Authors: Xiangxiong Kong
Abstract: Historic structures are important for our society but could be prone to structural deterioration due to long service durations and natural impacts. Monitoring the deterioration of historic structures becomes essential for stakeholders to take appropriate interventions. Existing work in the literature primarily focuses on assessing the structural damage at a given moment instead of evaluating the development of deterioration over time. To address this gap, we proposed a novel five-component digital twin framework to monitor time-varying changes in historic structures. A testbed of a casemate in Fort Soledad on the island of Guam was selected to validate our framework. Using this testbed, key implementation steps in our digital twin framework were performed. The findings from this study confirm that our digital twin framework can effectively monitor deterioration over time, which is an urgent need in the cultural heritage preservation community.
Authors: Haoran Zhang, Aparna Balagopalan, Nassim Oufattole, Hyewon Jeong, Yan Wu, Jiacheng Zhu, Marzyeh Ghassemi
Abstract: Large repositories of image-caption pairs are essential for the development of vision-language models. However, these datasets are often extracted from noisy data scraped from the web, and contain many mislabeled examples. In order to improve the reliability of downstream models, it is important to identify and filter images with incorrect captions. However, beyond filtering based on image-caption embedding similarity, no prior works have proposed other methods to filter noisy multimodal data, or concretely assessed the impact of noisy captioning data on downstream training. In this work, we propose LEMoN, a method to automatically identify label errors in multimodal datasets. Our method leverages the multimodal neighborhood of image-caption pairs in the latent space of contrastively pretrained multimodal models. We find that our method outperforms the baselines in label error identification, and that training on datasets filtered using our method improves downstream classification and captioning performance.
Authors: Peizhuo Li, Sebastian Starke, Yuting Ye, Olga Sorkine-Hornung
Abstract: We present a new approach for understanding the periodicity structure and semantics of motion datasets, independently of the morphology and skeletal structure of characters. Unlike existing methods using an overly sparse high-dimensional latent, we propose a phase manifold consisting of multiple closed curves, each corresponding to a latent amplitude. With our proposed vector quantized periodic autoencoder, we learn a shared phase manifold for multiple characters, such as a human and a dog, without any supervision. This is achieved by exploiting the discrete structure and a shallow network as bottlenecks, such that semantically similar motions are clustered into the same curve of the manifold, and the motions within the same component are aligned temporally by the phase variable. In combination with an improved motion matching framework, we demonstrate the manifold's capability of timing and semantics alignment in several applications, including motion retrieval, transfer and stylization. Code and pre-trained models for this paper are available at https://peizhuoli.github.io/walkthedog.
Authors: Stanley Cao, Sonny Young
Abstract: Image captioning using Vision Transformers (ViTs) represents a pivotal convergence of computer vision and natural language processing, offering the potential to enhance user experiences, improve accessibility, and provide textual representations of visual data. This paper explores the application of image captioning techniques to New Yorker cartoons, aiming to generate captions that emulate the wit and humor of winning entries in the New Yorker Cartoon Caption Contest. This task necessitates sophisticated visual and linguistic processing, along with an understanding of cultural nuances and humor. We propose several new baselines for using vision transformer encoder-decoder models to generate captions for the New Yorker cartoon caption contest.
Authors: Ahmed Alhamadah, Muntasir Mamun, Henry Harms, Mathew Redondo, Yu-Zheng Lin, Jesus Pacheco, Soheil Salehi, Pratik Satam
Abstract: The onset of Industry 4.0 is rapidly transforming the manufacturing world through the integration of cloud computing, machine learning (ML), artificial intelligence (AI), and universal network connectivity, resulting in performance optimization and increase productivity. Digital Twins (DT) are one such transformational technology that leverages software systems to replicate physical process behavior, representing the physical process in a digital environment. This paper aims to explore the use of photogrammetry (which is the process of reconstructing physical objects into virtual 3D models using photographs) and 3D Scanning techniques to create accurate visual representation of the 'Physical Process', to interact with the ML/AI based behavior models. To achieve this, we have used a readily available consumer device, the iPhone 15 Pro, which features stereo vision capabilities, to capture the depth of an Industry 4.0 system. By processing these images using 3D scanning tools, we created a raw 3D model for 3D modeling and rendering software for the creation of a DT model. The paper highlights the reliability of this method by measuring the error rate in between the ground truth (measurements done manually using a tape measure) and the final 3D model created using this method. The overall mean error is 4.97\% and the overall standard deviation error is 5.54\% between the ground truth measurements and their photogrammetry counterparts. The results from this work indicate that photogrammetry using consumer-grade devices can be an efficient and cost-efficient approach to creating DTs for smart manufacturing, while the approaches flexibility allows for iterative improvements of the models over time.
Authors: Xiaokai Chen, Xuan Liu, Donglin Di, Yongjia Ma, Wei Chen, Tonghua Su
Abstract: In recent years, facial video generation models have gained popularity. However, these models often lack expressive power when dealing with exaggerated anime-style faces due to the absence of high-quality anime-style face training sets. We propose a facial animation platform that enables real-time conversion from real human faces to cartoon-style faces, supporting multiple models. Built on the Gradio framework, our platform ensures excellent interactivity and user-friendliness. Users can input a real face video or image and select their desired cartoon style. The system will then automatically analyze facial features, execute necessary preprocessing, and invoke appropriate models to generate expressive anime-style faces. We employ a variety of models within our system to process the HDTF dataset, thereby creating an animated facial video dataset.
Authors: Syed Javed, Tariq M. Khan, Abdul Qayyum, Arcot Sowmya, Imran Razzak
Abstract: Retinal imaging has emerged as a promising method of addressing this challenge, taking advantage of the unique structure of the retina. The retina is an embryonic extension of the central nervous system, providing a direct in vivo window into neurological health. Recent studies have shown that specific structural changes in retinal vessels can not only serve as early indicators of various diseases but also help to understand disease progression. In this work, we present a lightweight retinal vessel segmentation network based on the encoder-decoder mechanism with region-guided attention. We introduce inverse addition attention blocks with region guided attention to focus on the foreground regions and improve the segmentation of regions of interest. To further boost the model's performance on retinal vessel segmentation, we employ a weighted dice loss. This choice is particularly effective in addressing the class imbalance issues frequently encountered in retinal vessel segmentation tasks. Dice loss penalises false positives and false negatives equally, encouraging the model to generate more accurate segmentation with improved object boundary delineation and reduced fragmentation. Extensive experiments on a benchmark dataset show better performance (0.8285, 0.8098, 0.9677, and 0.8166 recall, precision, accuracy and F1 score respectively) compared to state-of-the-art methods.
Authors: Baao Xie, Qiuyu Chen, Yunnan Wang, Zequn Zhang, Xin Jin, Wenjun Zeng
Abstract: Disentangled representation learning (DRL) aims to identify and decompose underlying factors behind observations, thus facilitating data perception and generation. However, current DRL approaches often rely on the unrealistic assumption that semantic factors are statistically independent. In reality, these factors may exhibit correlations, which off-the-shelf solutions have yet to properly address. To tackle this challenge, we introduce a bidirectional weighted graph-based framework, to learn factorized attributes and their interrelations within complex data. Specifically, we propose a $\beta$-VAE based module to extract factors as the initial nodes of the graph, and leverage the multimodal large language model (MLLM) to discover and rank latent correlations, thereby updating the weighted edges. By integrating these complementary modules, our model successfully achieves fine-grained, practical and unsupervised disentanglement. Experiments demonstrate our method's superior performance in disentanglement and reconstruction. Furthermore, the model inherits enhanced interpretability and generalizability from MLLMs.
Authors: Fernando Julio Cendra, Bingchen Zhao, Kai Han
Abstract: We tackle the problem of Continual Category Discovery (CCD), which aims to automatically discover novel categories in a continuous stream of unlabeled data while mitigating the challenge of catastrophic forgetting -- an open problem that persists even in conventional, fully supervised continual learning. To address this challenge, we propose PromptCCD, a simple yet effective framework that utilizes a Gaussian Mixture Model (GMM) as a prompting method for CCD. At the core of PromptCCD lies the Gaussian Mixture Prompting (GMP) module, which acts as a dynamic pool that updates over time to facilitate representation learning and prevent forgetting during category discovery. Moreover, GMP enables on-the-fly estimation of category numbers, allowing PromptCCD to discover categories in unlabeled data without prior knowledge of the category numbers. We extend the standard evaluation metric for Generalized Category Discovery (GCD) to CCD and benchmark state-of-the-art methods on diverse public datasets. PromptCCD significantly outperforms existing methods, demonstrating its effectiveness. Project page: https://visual-ai.github.io/promptccd .
Authors: Zhijian Liu, Zhuoyang Zhang, Samir Khaki, Shang Yang, Haotian Tang, Chenfeng Xu, Kurt Keutzer, Song Han
Abstract: Semantic segmentation empowers numerous real-world applications, such as autonomous driving and augmented/mixed reality. These applications often operate on high-resolution images (e.g., 8 megapixels) to capture the fine details. However, this comes at the cost of considerable computational complexity, hindering the deployment in latency-sensitive scenarios. In this paper, we introduce SparseRefine, a novel approach that enhances dense low-resolution predictions with sparse high-resolution refinements. Based on coarse low-resolution outputs, SparseRefine first uses an entropy selector to identify a sparse set of pixels with high entropy. It then employs a sparse feature extractor to efficiently generate the refinements for those pixels of interest. Finally, it leverages a gated ensembler to apply these sparse refinements to the initial coarse predictions. SparseRefine can be seamlessly integrated into any existing semantic segmentation model, regardless of CNN- or ViT-based. SparseRefine achieves significant speedup: 1.5 to 3.7 times when applied to HRNet-W48, SegFormer-B5, Mask2Former-T/L and SegNeXt-L on Cityscapes, with negligible to no loss of accuracy. Our "dense+sparse" paradigm paves the way for efficient high-resolution visual computing.
Authors: Hikaru Ikuta, Leslie W\"ohler, Kiyoharu Aizawa
Abstract: Manga is a popular medium that combines stylized drawings and text to convey stories. As manga panels differ from natural images, computational systems traditionally had to be designed specifically for manga. Recently, the adaptive nature of modern large multimodal models (LMMs) shows possibilities for more general approaches. To provide an analysis of the current capability of LMMs for manga understanding tasks and identifying areas for their improvement, we design and evaluate MangaUB, a novel manga understanding benchmark for LMMs. MangaUB is designed to assess the recognition and understanding of content shown in a single panel as well as conveyed across multiple panels, allowing for a fine-grained analysis of a model's various capabilities required for manga understanding. Our results show strong performance on the recognition of image content, while understanding the emotion and information conveyed across multiple panels is still challenging, highlighting future work towards LMMs for manga understanding.
Authors: Shen Chen, Jiale Zhou, Zhongyu Jiang, Tianfang Zhang, Zongkai Wu, Jenq-Neng Hwang, Lei Li
Abstract: The creation of high-quality 3D assets is paramount for applications in digital heritage preservation, entertainment, and robotics. Traditionally, this process necessitates skilled professionals and specialized software for the modeling, texturing, and rendering of 3D objects. However, the rising demand for 3D assets in gaming and virtual reality (VR) has led to the creation of accessible image-to-3D technologies, allowing non-professionals to produce 3D content and decreasing dependence on expert input. Existing methods for 3D content generation struggle to simultaneously achieve detailed textures and strong geometric consistency. We introduce a novel 3D content creation framework, ScalingGaussian, which combines 3D and 2D diffusion models to achieve detailed textures and geometric consistency in generated 3D assets. Initially, a 3D diffusion model generates point clouds, which are then densified through a process of selecting local regions, introducing Gaussian noise, followed by using local density-weighted selection. To refine the 3D gaussians, we utilize a 2D diffusion model with Score Distillation Sampling (SDS) loss, guiding the 3D Gaussians to clone and split. Finally, the 3D Gaussians are converted into meshes, and the surface textures are optimized using Mean Square Error(MSE) and Gradient Profile Prior(GPP) losses. Our method addresses the common issue of sparse point clouds in 3D diffusion, resulting in improved geometric structure and detailed textures. Experiments on image-to-3D tasks demonstrate that our approach efficiently generates high-quality 3D assets.
Authors: Hojin Jang, Pawan Sinha, Xavier Boix
Abstract: Configural processing, the perception of spatial relationships among an object's components, is crucial for object recognition. However, the teleology and underlying neurocomputational mechanisms of such processing are still elusive, notwithstanding decades of research. We hypothesized that processing objects via configural cues provides a more robust means to recognizing them relative to local featural cues. We evaluated this hypothesis by devising identification tasks with composite letter stimuli and comparing different neural network models trained with either only local or configural cues available. We found that configural cues yielded more robust performance to geometric transformations such as rotation or scaling. Furthermore, when both features were simultaneously available, configural cues were favored over local featural cues. Layerwise analysis revealed that the sensitivity to configural cues emerged later relative to local feature cues, possibly contributing to the robustness to pixel-level transformations. Notably, this configural processing occurred in a purely feedforward manner, without the need for recurrent computations. Our findings with letter stimuli were successfully extended to naturalistic face images. Thus, our study provides neurocomputational evidence that configural processing emerges in a na\"ive network based on task contingencies, and is beneficial for robust object processing under varying viewing conditions.
Authors: Abu Taib Mohammed Shahjahan, A. Ben Hamza
Abstract: Although graph convolutional networks exhibit promising performance in 3D human pose estimation, their reliance on one-hop neighbors limits their ability to capture high-order dependencies among body joints, crucial for mitigating uncertainty arising from occlusion or depth ambiguity. To tackle this limitation, we introduce Flex-GCN, a flexible graph convolutional network designed to learn graph representations that capture broader global information and dependencies. At its core is the flexible graph convolution, which aggregates features from both immediate and second-order neighbors of each node, while maintaining the same time and memory complexity as the standard convolution. Our network architecture comprises residual blocks of flexible graph convolutional layers, as well as a global response normalization layer for global feature aggregation, normalization and calibration. Quantitative and qualitative results demonstrate the effectiveness of our model, achieving competitive performance on benchmark datasets.
Authors: Ziyuan Fang, Hanqing Zhao, Tianyi Wei, Wenbo Zhou, Ming Wan, Zhanyi Wang, Weiming Zhang, Nenghai Yu
Abstract: Previous deepfake detection methods mostly depend on low-level textural features vulnerable to perturbations and fall short of detecting unseen forgery methods. In contrast, high-level semantic features are less susceptible to perturbations and not limited to forgery-specific artifacts, thus having stronger generalization. Motivated by this, we propose a detection method that utilizes high-level semantic features of faces to identify inconsistencies in temporal domain. We introduce UniForensics, a novel deepfake detection framework that leverages a transformer-based video classification network, initialized with a meta-functional face encoder for enriched facial representation. In this way, we can take advantage of both the powerful spatio-temporal model and the high-level semantic information of faces. Furthermore, to leverage easily accessible real face data and guide the model in focusing on spatio-temporal features, we design a Dynamic Video Self-Blending (DVSB) method to efficiently generate training samples with diverse spatio-temporal forgery traces using real facial videos. Based on this, we advance our framework with a two-stage training approach: The first stage employs a novel self-supervised contrastive learning, where we encourage the network to focus on forgery traces by impelling videos generated by the same forgery process to have similar representations. On the basis of the representation learned in the first stage, the second stage involves fine-tuning on face forgery detection dataset to build a deepfake detector. Extensive experiments validates that UniForensics outperforms existing face forgery methods in generalization ability and robustness. In particular, our method achieves 95.3\% and 77.2\% cross dataset AUC on the challenging Celeb-DFv2 and DFDC respectively.
Authors: Gemmechu Hassena, Jonathan Moon, Ryan Fujii, Andrew Yuen, Noah Snavely, Steve Marschner, Bharath Hariharan
Abstract: Implicit neural fields have made remarkable progress in reconstructing 3D surfaces from multiple images; however, they encounter challenges when it comes to separating individual objects within a scene. Previous work has attempted to tackle this problem by introducing a framework to train separate signed distance fields (SDFs) simultaneously for each of N objects and using a regularization term to prevent objects from overlapping. However, all of these methods require segmentation masks to be provided, which are not always readily available. We introduce our method, ObjectCarver, to tackle the problem of object separation from just click input in a single view. Given posed multi-view images and a set of user-input clicks to prompt segmentation of the individual objects, our method decomposes the scene into separate objects and reconstructs a high-quality 3D surface for each one. We introduce a loss function that prevents floaters and avoids inappropriate carving-out due to occlusion. In addition, we introduce a novel scene initialization method that significantly speeds up the process while preserving geometric details compared to previous approaches. Despite requiring neither ground truth masks nor monocular cues, our method outperforms baselines both qualitatively and quantitatively. In addition, we introduce a new benchmark dataset for evaluation.
Authors: Xiaoyan Yu, Shen Zhou, Huafeng Li, Liehuang Zhu
Abstract: The use of a single image restoration framework to achieve multi-task image restoration has garnered significant attention from researchers. However, several practical challenges remain, including meeting the specific and simultaneous demands of different tasks, balancing relationships between tasks, and effectively utilizing task correlations in model design. To address these challenges, this paper explores a multi-expert adaptive selection mechanism. We begin by designing a feature representation method that accounts for both the pixel channel level and the global level, encompassing low-frequency and high-frequency components of the image. Based on this method, we construct a multi-expert selection and ensemble scheme. This scheme adaptively selects the most suitable expert from the expert library according to the content of the input image and the prompts of the current task. It not only meets the individualized needs of different tasks but also achieves balance and optimization across tasks. By sharing experts, our design promotes interconnections between different tasks, thereby enhancing overall performance and resource utilization. Additionally, the multi-expert mechanism effectively eliminates irrelevant experts, reducing interference from them and further improving the effectiveness and accuracy of image restoration. Experimental results demonstrate that our proposed method is both effective and superior to existing approaches, highlighting its potential for practical applications in multi-task image restoration.
Authors: Xiaoxiao Wu, Xiaowei Chen, Zhenguo Gao, Shulei Qu, Yuanyuan Qiu
Abstract: Medical image segmentation has witnessed significant advancements with the emergence of deep learning. However, the reliance of most neural network models on a substantial amount of annotated data remains a challenge for medical image segmentation. To address this issue, few-shot segmentation methods based on meta-learning have been employed. Presently, the methods primarily focus on aligning the support set and query set to enhance performance, but this approach hinders further improvement of the model's effectiveness. In this paper, our objective is to propose a few-shot medical segmentation model that acquire comprehensive feature representation capabilities, which will boost segmentation accuracy by capturing both local and long-range features. To achieve this, we introduce a plug-and-play attention module that dynamically enhances both query and support features, thereby improving the representativeness of the extracted features. Our model comprises four key modules: a dual-path feature extractor, an attention module, an adaptive prototype prediction module, and a multi-scale prediction fusion module. Specifically, the dual-path feature extractor acquires multi-scale features by obtaining features of 32{\times}32 size and 64{\times}64 size. The attention module follows the feature extractor and captures local and long-range information. The adaptive prototype prediction module automatically adjusts the anomaly score threshold to predict prototypes, while the multi-scale fusion prediction module integrates prediction masks of various scales to produce the final segmentation result. We conducted experiments on publicly available MRI datasets, namely CHAOS and CMR, and compared our method with other advanced techniques. The results demonstrate that our method achieves state-of-the-art performance.
Authors: Shengjie Zhu, Girish Chandar Ganesan, Abhinav Kumar, Xiaoming Liu
Abstract: 3D sensing is a fundamental task for Autonomous Vehicles. Its deployment often relies on aligned RGB cameras and LiDAR. Despite meticulous synchronization and calibration, systematic misalignment persists in LiDAR projected depthmap. This is due to the physical baseline distance between the two sensors. The artifact is often reflected as background LiDAR incorrectly projected onto the foreground, such as cars and pedestrians. The KITTI dataset uses stereo cameras as a heuristic solution to remove artifacts. However most AV datasets, including nuScenes, Waymo, and DDAD, lack stereo images, making the KITTI solution inapplicable. We propose RePLAy, a parameter-free analytical solution to remove the projective artifacts. We construct a binocular vision system between a hypothesized virtual LiDAR camera and the RGB camera. We then remove the projective artifacts by determining the epipolar occlusion with the proposed analytical solution. We show unanimous improvement in the State-of-The-Art (SoTA) monocular depth estimators and 3D object detectors with the artifacts-free depthmaps.
Authors: Juhan Cha, Minseok Joo, Jihwan Park, Sanghyeok Lee, Injae Kim, Hyunwoo J. Kim
Abstract: Recent advancements in 3D object detection have benefited from multi-modal information from the multi-view cameras and LiDAR sensors. However, the inherent disparities between the modalities pose substantial challenges. We observe that existing multi-modal 3D object detection methods heavily rely on the LiDAR sensor, treating the camera as an auxiliary modality for augmenting semantic details. This often leads to not only underutilization of camera data but also significant performance degradation in scenarios where LiDAR data is unavailable. Additionally, existing fusion methods overlook the detrimental impact of sensor noise induced by environmental changes, on detection performance. In this paper, we propose MEFormer to address the LiDAR over-reliance problem by harnessing critical information for 3D object detection from every available modality while concurrently safeguarding against corrupted signals during the fusion process. Specifically, we introduce Modality Agnostic Decoding (MOAD) that extracts geometric and semantic features with a shared transformer decoder regardless of input modalities and provides promising improvement with a single modality as well as multi-modality. Additionally, our Proximity-based Modality Ensemble (PME) module adaptively utilizes the strengths of each modality depending on the environment while mitigating the effects of a noisy sensor. Our MEFormer achieves state-of-the-art performance of 73.9% NDS and 71.5% mAP in the nuScenes validation set. Extensive analyses validate that our MEFormer improves robustness against challenging conditions such as sensor malfunctions or environmental changes. The source code is available at https://github.com/hanchaa/MEFormer
Authors: Shengjie Zhu, Xiaoming Liu
Abstract: Both self-supervised depth estimation and Structure-from-Motion (SfM) recover scene depth from RGB videos. Despite sharing a similar objective, the two approaches are disconnected. Prior works of self-supervision backpropagate losses defined within immediate neighboring frames. Instead of learning-through-loss, this work proposes an alternative scheme by performing local SfM. First, with calibrated RGB or RGB-D images, we employ a depth and correspondence estimator to infer depthmaps and pair-wise correspondence maps. Then, a novel bundle-RANSAC-adjustment algorithm jointly optimizes camera poses and one depth adjustment for each depthmap. Finally, we fix camera poses and employ a NeRF, however, without a neural network, for dense triangulation and geometric verification. Poses, depth adjustments, and triangulated sparse depths are our outputs. For the first time, we show self-supervision within $5$ frames already benefits SoTA supervised depth and correspondence models.
Authors: Shuran Ma, Weiying Xie, Daixun Li, Haowei Li, Yunsong Li
Abstract: The rapid development of multimedia has provided a large amount of data with different distributions for visual tasks, forming different domains. Federated Learning (FL) can efficiently use this diverse data distributed on different client media in a decentralized manner through model sharing. However, in open-world scenarios, there is a challenge: global models may struggle to predict well on entirely new domain data captured by certain media, which were not encountered during training. Existing methods still rely on strong statistical correlations between samples and labels to address this issue, which can be misleading, as some features may establish spurious short-cut correlations with the predictions. To comprehensively address this challenge, we introduce FedCD (Cross-Domain Invariant Federated Learning), an overall optimization framework at both the local and global levels. We introduce the Spurious Correlation Intervener (SCI), which employs invariance theory to locally generate interventers for features in a self-supervised manner to reduce the model's susceptibility to spurious correlated features. Our approach requires no sharing of data or features, only the gradients related to the model. Additionally, we develop the simple yet effective Risk Extrapolation Aggregation strategy (REA), determining aggregation coefficients through mathematical optimization to facilitate global causal invariant predictions. Extensive experiments and ablation studies highlight the effectiveness of our approach. In both classification and object detection generalization tasks, our method outperforms the baselines by an average of at least 1.45% in Acc, 4.8% and 1.27% in mAP50.
Authors: Jiahao Wang, Mingxuan Li, Haichen Luo, Jinguo Zhu, Aijun Yang, Mingzhe Rong, Xiaohua Wang
Abstract: The inspection of power transmission line has achieved notable achievements in the past few years, primarily due to the integration of deep learning technology. However, current inspection approaches continue to encounter difficulties in generalization and intelligence, which restricts their further applicability. In this paper, we introduce Power-LLaVA, the first large language and vision assistant designed to offer professional and reliable inspection services for power transmission line by engaging in dialogues with humans. Moreover, we also construct a large-scale and high-quality dataset specialized for the inspection task. By employing a two-stage training strategy on the constructed dataset, Power-LLaVA demonstrates exceptional performance at a comparatively low training cost. Extensive experiments further prove the great capabilities of Power-LLaVA within the realm of power transmission line inspection. Code shall be released.
Authors: Yinheng Li, Han Ding, Hang Chen
Abstract: Data processing plays an significant role in current multimodal model training. In this paper. we provide an comprehensive review of common data processing techniques used in modern multimodal model training with a focus on diffusion models and multimodal large language models (MLLMs). We summarized all techniques into four categories: data quality, data quantity, data distribution and data safety. We further present our findings in the choice of data process methods in different type of models. This study aims to provide guidance to multimodal models developers with effective data processing techniques.
Authors: Jinda Zhang, Michal Aibin
Abstract: Forest fires pose a significant threat to ecosystems, economies, and human health worldwide. Early detection and assessment of forest fires are crucial for effective management and conservation efforts. Unmanned Aerial Vehicles (UAVs) equipped with advanced computer vision algorithms offer a promising solution for forest fire detection and assessment. In this paper, we optimize an integrated forest fire risk assessment framework using UAVs and multi-stage object detection algorithms. We introduce improvements to our previous framework, including the adoption of Faster R-CNN, Grid R-CNN, Sparse R-CNN, Cascade R-CNN, Dynamic R-CNN, and Libra R-CNN detectors, and explore optimizations such as CBAM for attention enhancement, random erasing for preprocessing, and different color space representations. We evaluate these enhancements through extensive experimentation using aerial image footage from various regions in British Columbia, Canada. Our findings demonstrate the effectiveness of multi-stage detectors and optimizations in improving the accuracy of forest fire risk assessment. This research contributes to the advancement of UAV-based forest fire detection and assessment systems, enhancing their efficiency and effectiveness in supporting sustainable forest management and conservation efforts.
Authors: Ruiyi Zhang, Yufan Zhou, Jian Chen, Jiuxiang Gu, Changyou Chen, Tong Sun
Abstract: Large multimodal language models have demonstrated impressive capabilities in understanding and manipulating images. However, many of these models struggle with comprehending intensive textual contents embedded within the images, primarily due to the limited text recognition and layout understanding ability. To understand the sources of these limitations, we perform an exploratory analysis showing the drawbacks of classical visual encoders on visual text understanding. Hence, we present LLaVA-Read, a multimodal large language model that utilizes dual visual encoders along with a visual text encoder. Our model surpasses existing state-of-the-art models in various text-rich image understanding tasks, showcasing enhanced comprehension of textual content within images. Together, our research suggests visual text understanding remains an open challenge and an efficient visual text encoder is crucial for future successful multimodal systems.
Authors: Ashkan Taghipour, Morteza Ghahremani, Mohammed Bennamoun, Aref Miri Rekavandi, Zinuo Li, Hamid Laga, Farid Boussaid
Abstract: This paper investigates the role of CLIP image embeddings within the Stable Video Diffusion (SVD) framework, focusing on their impact on video generation quality and computational efficiency. Our findings indicate that CLIP embeddings, while crucial for aesthetic quality, do not significantly contribute towards the subject and background consistency of video outputs. Moreover, the computationally expensive cross-attention mechanism can be effectively replaced by a simpler linear layer. This layer is computed only once at the first diffusion inference step, and its output is then cached and reused throughout the inference process, thereby enhancing efficiency while maintaining high-quality outputs. Building on these insights, we introduce the VCUT, a training-free approach optimized for efficiency within the SVD architecture. VCUT eliminates temporal cross-attention and replaces spatial cross-attention with a one-time computed linear layer, significantly reducing computational load. The implementation of VCUT leads to a reduction of up to 322T Multiple-Accumulate Operations (MACs) per video and a decrease in model parameters by up to 50M, achieving a 20% reduction in latency compared to the baseline. Our approach demonstrates that conditioning during the Semantic Binding stage is sufficient, eliminating the need for continuous computation across all inference steps and setting a new standard for efficient video generation.
Authors: Ying Zang, Yidong Han, Chaotao Ding, Jianqi Zhang, Tianrun Chen
Abstract: The requirement for 3D content is growing as AR/VR application emerges. At the same time, 3D modelling is only available for skillful experts, because traditional methods like Computer-Aided Design (CAD) are often too labor-intensive and skill-demanding, making it challenging for novice users. Our proposed method, Magic3DSketch, employs a novel technique that encodes sketches to predict a 3D mesh, guided by text descriptions and leveraging external prior knowledge obtained through text and language-image pre-training. The integration of language-image pre-trained neural networks complements the sparse and ambiguous nature of single-view sketch inputs. Our method is also more useful and offers higher degree of controllability compared to existing text-to-3D approaches, according to our user study. Moreover, Magic3DSketch achieves state-of-the-art performance in both synthetic and real dataset with the capability of producing more detailed structures and realistic shapes with the help of text input. Users are also more satisfied with models obtained by Magic3DSketch according to our user study. Additionally, we are also the first, to our knowledge, add color based on text description to the sketch-derived shapes. By combining sketches and text guidance with the help of language-image pretrained models, our Magic3DSketch can allow novice users to create custom 3D models with minimal effort and maximum creative freedom, with the potential to revolutionize future 3D modeling pipelines.
Authors: Penghui Wen, Kun Hu, Dong Yuan, Zhiyuan Ning, Changyang Li, Zhiyong Wang
Abstract: Radio frequency (RF) signals have been proved to be flexible for human silhouette segmentation (HSS) under complex environments. Existing studies are mainly based on a one-shot approach, which lacks a coherent projection ability from the RF domain. Additionally, the spatio-temporal patterns have not been fully explored for human motion dynamics in HSS. Therefore, we propose a two-stage Sequential Diffusion Model (SDM) to progressively synthesize high-quality segmentation jointly with the considerations on motion dynamics. Cross-view transformation blocks are devised to guide the diffusion model in a multi-scale manner for comprehensively characterizing human related patterns in an individual frame such as directional projection from signal planes. Moreover, spatio-temporal blocks are devised to fine-tune the frame-level model to incorporate spatio-temporal contexts and motion dynamics, enhancing the consistency of the segmentation maps. Comprehensive experiments on a public benchmark -- HIBER demonstrate the state-of-the-art performance of our method with an IoU 0.732. Our code is available at https://github.com/ph-w2000/SDM.
Authors: Yansheng Li, Tingzhu Wang, Kang Wu, Linlin Wang, Xin Guo, Wenbin Wang
Abstract: Scene Graph Generation (SGG) aims to explore the relationships between objects in images and obtain scene summary graphs, thereby better serving downstream tasks. However, the long-tailed problem has adversely affected the scene graph's quality. The predictions are dominated by coarse-grained relationships, lacking more informative fine-grained ones. The union region of one object pair (i.e., one sample) contains rich and dedicated contextual information, enabling the prediction of the sample-specific bias for refining the original relationship prediction. Therefore, we propose a novel Sample-Level Bias Prediction (SBP) method for fine-grained SGG (SBG). Firstly, we train a classic SGG model and construct a correction bias set by calculating the margin between the ground truth label and the predicted label with one classic SGG model. Then, we devise a Bias-Oriented Generative Adversarial Network (BGAN) that learns to predict the constructed correction biases, which can be utilized to correct the original predictions from coarse-grained relationships to fine-grained ones. The extensive experimental results on VG, GQA, and VG-1800 datasets demonstrate that our SBG outperforms the state-of-the-art methods in terms of Average@K across three mainstream SGG models: Motif, VCtree, and Transformer. Compared to dataset-level correction methods on VG, SBG shows a significant average improvement of 5.6%, 3.9%, and 3.2% on Average@K for tasks PredCls, SGCls, and SGDet, respectively. The code will be available at https://github.com/Zhuzi24/SBG.
Authors: Gang Pan, Chen Wang, Zhijie Sui, Shuai Guo, Yaozhi Lv, Honglie Li, Di Sun
Abstract: The Quick-view (QV) technique serves as a primary method for detecting defects within sewerage systems. However, the effectiveness of QV is impeded by the limited visual range of its hardware, resulting in suboptimal image quality for distant portions of the sewer network. Image super-resolution is an effective way to improve image quality and has been applied in a variety of scenes. However, research on super-resolution for sewer images remains considerably unexplored. In response, this study leverages the inherent depth relationships present within QV images and introduces a novel Depth-guided, Reference-based Super-Resolution framework denoted as DSRNet. It comprises two core components: a depth extraction module and a depth information matching module (DMM). DSRNet utilizes the adjacent frames of the low-resolution image as reference images and helps them recover texture information based on the correlation. By combining these modules, the integration of depth priors significantly enhances both visual quality and performance benchmarks. Besides, in pursuit of computational efficiency and compactness, our paper introduces a super-resolution knowledge distillation model based on an attention mechanism. This mechanism facilitates the acquisition of feature similarity between a more complex teacher model and a streamlined student model, the latter being a lightweight version of DSRNet. Experimental results demonstrate that DSRNet significantly improves PSNR and SSIM compared with other methods. This study also conducts experiments on sewer defect semantic segmentation, object detection, and classification on the Pipe dataset and Sewer-ML dataset. Experiments show that the method can improve the performance of low-resolution sewer images in these tasks.
Authors: Bailiang Jian, Jiazhen Pan, Morteza Ghahremani, Daniel Rueckert, Christian Wachinger, Benedikt Wiestler
Abstract: Our findings indicate that adopting "advanced" computational elements fails to significantly improve registration accuracy. Instead, well-established registration-specific designs offer fair improvements, enhancing results by a marginal 1.5\% over the baseline. Our findings emphasize the importance of rigorous, unbiased evaluation and contribution disentanglement of all low- and high-level registration components, rather than simply following the computer vision trends with "more advanced" computational blocks. We advocate for simpler yet effective solutions and novel evaluation metrics that go beyond conventional registration accuracy, warranting further research across diverse organs and modalities. The code is available at \url{https://github.com/BailiangJ/rethink-reg}.
Authors: Chengzhi Wu, Kaige Wang, Zeyun Zhong, Hao Fu, Junwei Zheng, Jiaming Zhang, Julius Pfrommer, J\"urgen Beyerer
Abstract: In recent years, there have been significant advancements in applying attention mechanisms to point cloud analysis. However, attention module variants featured in various research papers often operate under diverse settings and tasks, incorporating potential training strategies. This heterogeneity poses challenges in establishing a fair comparison among these attention module variants. In this paper, we address this issue by rethinking and exploring attention module design within a consistent base framework and settings. Both global-based and local-based attention methods are studied, with a focus on the selection basis and scales of neighbors for local-based attention. Different combinations of aggregated local features and computation methods for attention scores are evaluated, ranging from the initial addition/concatenation-based approach to the widely adopted dot product-based method and the recently proposed vector attention technique. Various position encoding methods are also investigated. Our extensive experimental analysis reveals that there is no universally optimal design across diverse point cloud tasks. Instead, drawing from best practices, we propose tailored attention modules for specific tasks, leading to superior performance on point cloud classification and segmentation benchmarks.
Authors: Samuel Schmidgall, Joseph Cho, Cyril Zakka, William Hiesinger
Abstract: Surgery requires comprehensive medical knowledge, visual assessment skills, and procedural expertise. While recent surgical AI models have focused on solving task-specific problems, there is a need for general-purpose systems that can understand surgical scenes and interact through natural language. This paper introduces GP-VLS, a general-purpose vision language model for surgery that integrates medical and surgical knowledge with visual scene understanding. For comprehensively evaluating general-purpose surgical models, we propose SurgiQual, which evaluates across medical and surgical knowledge benchmarks as well as surgical vision-language questions. To train GP-VLS, we develop six new datasets spanning medical knowledge, surgical textbooks, and vision-language pairs for tasks like phase recognition and tool identification. We show that GP-VLS significantly outperforms existing open- and closed-source models on surgical vision-language tasks, with 8-21% improvements in accuracy across SurgiQual benchmarks. GP-VLS also demonstrates strong performance on medical and surgical knowledge tests compared to open-source alternatives. Overall, GP-VLS provides an open-source foundation for developing AI assistants to support surgeons across a wide range of tasks and scenarios.
Authors: Qun Li, Baoquan Sun, Fu Xiao, Yonggang Qi, Bir Bhanu
Abstract: We propose Sym-Net, a novel framework for Few-Shot Segmentation (FSS) that addresses the critical issue of intra-class variation by jointly learning both query and support prototypes in a symmetrical manner. Unlike previous methods that generate query prototypes solely by matching query features to support prototypes, which is a form of bias learning towards the few-shot support samples, Sym-Net leverages a balanced symmetrical learning approach for both query and support prototypes, ensuring that the learning process does not favor one set (support or query) over the other. One of main modules of Sym-Net is the visual-text alignment-based prototype aggregation module, which is not just query-guided prototype refinement, it is a jointly learning from both support and query samples, which makes the model beneficial for handling intra-class discrepancies and allows it to generalize better to new, unseen classes. Specifically, a parameter-free prior mask generation module is designed to accurately localize both local and global regions of the query object by using sliding windows of different sizes and a self-activation kernel to suppress incorrect background matches. Additionally, to address the information loss caused by spatial pooling during prototype learning, a top-down hyper-correlation module is integrated to capture multi-scale spatial relationships between support and query images. This approach is further jointly optimized by implementing a co-optimized hard triplet mining strategy. Experimental results show that the proposed Sym-Net outperforms state-of-the-art models, which demonstrates that jointly learning support-query prototypes in a symmetrical manner for FSS offers a promising direction to enhance segmentation performance with limited annotated data.
Authors: Xianren Zhang, Dongwon Lee, Suhang Wang
Abstract: As deep vision models' popularity rapidly increases, there is a growing emphasis on explanations for model predictions. The inherently explainable attribution method aims to enhance the understanding of model behavior by identifying the important regions in images that significantly contribute to predictions. It is achieved by cooperatively training a selector (generating an attribution map to identify important features) and a predictor (making predictions using the identified features). Despite many advancements, existing methods suffer from the incompleteness problem, where discriminative features are masked out, and the interlocking problem, where the non-optimized selector initially selects noise, causing the predictor to fit on this noise and perpetuate the cycle. To address these problems, we introduce a new objective that discourages the presence of discriminative features in the masked-out regions thus enhancing the comprehensiveness of feature selection. A pre-trained detector is introduced to detect discriminative features in the masked-out region. If the selector selects noise instead of discriminative features, the detector can observe and break the interlocking situation by penalizing the selector. Extensive experiments show that our model makes accurate predictions with higher accuracy than the regular black-box model, and produces attribution maps with high feature coverage, localization ability, fidelity and robustness. Our code will be available at \href{https://github.com/Zood123/COMET}{https://github.com/Zood123/COMET}.
URLs: https://github.com/Zood123/COMET, https://github.com/Zood123/COMET
Authors: Patryk Kuban, Michal Kawulok
Abstract: Detecting and segmenting human skin regions in digital images is an intensively explored topic of computer vision with a variety of approaches proposed over the years that have been found useful in numerous practical applications. The first methods were based on pixel-wise skin color modeling and they were later enhanced with context-based analysis to include the textural and geometrical features, recently extracted using deep convolutional neural networks. It has been also demonstrated that skin regions can be segmented from grayscale images without using color information at all. However, the possibility to combine these two sources of information has not been explored so far and we address this research gap with the contribution reported in this paper. We propose to train a convolutional network using the datasets focused on different features to create an ensemble whose individual outcomes are effectively combined using yet another convolutional network trained to produce the final segmentation map. The experimental results clearly indicate that the proposed approach outperforms the basic classifiers, as well as an ensemble based on the voting scheme. We expect that this study will help in developing new ensemble-based techniques that will improve the performance of semantic segmentation systems, reaching beyond the problem of detecting human skin.
Authors: Zhenlong Yuan, Cong Liu, Fei Shen, Zhaoxin Li, Tianlu Mao, Zhaoqi Wang
Abstract: Reconstructing textureless areas in MVS poses challenges due to the absence of reliable pixel correspondences within fixed patch. Although certain methods employ patch deformation to expand the receptive field, their patches mistakenly skip depth edges to calculate areas with depth discontinuity, thereby causing ambiguity. Consequently, we introduce Multi-granularity Segmentation Prior Multi-View Stereo (MSP-MVS). Specifically, we first propose multi-granularity segmentation prior by integrating multi-granularity depth edges to restrict patch deformation within homogeneous areas. Moreover, we present anchor equidistribution that bring deformed patches with more uniformly distributed anchors to ensure an adequate coverage of their own homogeneous areas. Furthermore, we introduce iterative local search optimization to represent larger patch with sparse representative candidates, significantly boosting the expressive capacity for each patch. The state-of-the-art results on ETH3D and Tanks & Temples benchmarks demonstrate the effectiveness and robust generalization ability of our proposed method.
Authors: Santosh V. Patapati
Abstract: Major Depressive Disorder (MDD) is a pervasive mental health condition that affects 300 million people worldwide. This work presents a novel, BiLSTM-based tri-modal model-level fusion architecture for the binary classification of depression from clinical interview recordings. The proposed architecture incorporates Mel Frequency Cepstral Coefficients, Facial Action Units, and uses a two-shot learning based GPT-4 model to process text data. This is the first work to incorporate large language models into a multi-modal architecture for this task. It achieves impressive results on the DAIC-WOZ AVEC 2016 Challenge cross-validation split and Leave-One-Subject-Out cross-validation split, surpassing all baseline models and multiple state-of-the-art models. In Leave-One-Subject-Out testing, it achieves an accuracy of 91.01%, an F1-Score of 85.95%, a precision of 80%, and a recall of 92.86%.
Authors: Peiming Li, Ziyi Wang, Mengyuan Liu, Hong Liu, Chen Chen
Abstract: Grasp generation aims to create complex hand-object interactions with a specified object. While traditional approaches for hand generation have primarily focused on visibility and diversity under scene constraints, they tend to overlook the fine-grained hand-object interactions such as contacts, resulting in inaccurate and undesired grasps. To address these challenges, we propose a controllable grasp generation task and introduce ClickDiff, a controllable conditional generation model that leverages a fine-grained Semantic Contact Map (SCM). Particularly when synthesizing interactive grasps, the method enables the precise control of grasp synthesis through either user-specified or algorithmically predicted Semantic Contact Map. Specifically, to optimally utilize contact supervision constraints and to accurately model the complex physical structure of hands, we propose a Dual Generation Framework. Within this framework, the Semantic Conditional Module generates reasonable contact maps based on fine-grained contact information, while the Contact Conditional Module utilizes contact maps alongside object point clouds to generate realistic grasps. We evaluate the evaluation criteria applicable to controllable grasp generation. Both unimanual and bimanual generation experiments on GRAB and ARCTIC datasets verify the validity of our proposed method, demonstrating the efficacy and robustness of ClickDiff, even with previously unseen objects. Our code is available at https://github.com/adventurer-w/ClickDiff.
Authors: Nagur Shareef Shaik, Teja Krishna Cherukuri, Vince D. Calhoun, Dong Hye Ye
Abstract: Schizophrenia (SZ) is a severe brain disorder marked by diverse cognitive impairments, abnormalities in brain structure, function, and genetic factors. Its complex symptoms and overlap with other psychiatric conditions challenge traditional diagnostic methods, necessitating advanced systems to improve precision. Existing research studies have mostly focused on imaging data, such as structural and functional MRI, for SZ diagnosis. There has been less focus on the integration of genomic features despite their potential in identifying heritable SZ traits. In this study, we introduce a Multi-modal Imaging Genomics Transformer (MIGTrans), that attentively integrates genomics with structural and functional imaging data to capture SZ-related neuroanatomical and connectome abnormalities. MIGTrans demonstrated improved SZ classification performance with an accuracy of 86.05% (+/- 0.02), offering clear interpretations and identifying significant genomic locations and brain morphological/connectivity patterns associated with SZ.
Authors: Tianxiao Zhang, Wenju Xu, Bo Luo, Guanghui Wang
Abstract: The Vision Transformer (ViT) leverages the Transformer's encoder to capture global information by dividing images into patches and achieves superior performance across various computer vision tasks. However, the self-attention mechanism of ViT captures the global context from the outset, overlooking the inherent relationships between neighboring pixels in images or videos. Transformers mainly focus on global information while ignoring the fine-grained local details. Consequently, ViT lacks inductive bias during image or video dataset training. In contrast, convolutional neural networks (CNNs), with their reliance on local filters, possess an inherent inductive bias, making them more efficient and quicker to converge than ViT with less data. In this paper, we present a lightweight Depth-Wise Convolution module as a shortcut in ViT models, bypassing entire Transformer blocks to ensure the models capture both local and global information with minimal overhead. Additionally, we introduce two architecture variants, allowing the Depth-Wise Convolution modules to be applied to multiple Transformer blocks for parameter savings, and incorporating independent parallel Depth-Wise Convolution modules with different kernels to enhance the acquisition of local information. The proposed approach significantly boosts the performance of ViT models on image classification, object detection and instance segmentation by a large margin, especially on small datasets, as evaluated on CIFAR-10, CIFAR-100, Tiny-ImageNet and ImageNet for image classification, and COCO for object detection and instance segmentation. The source code can be accessed at https://github.com/ZTX-100/Efficient_ViT_with_DW.
Authors: Haifeng Zhao, Lixiang Jiang, Leilei Ma, Dengdi Sun, Yanping Fu
Abstract: Medical images from different healthcare centers exhibit varied data distributions, posing significant challenges for adapting lung nodule detection due to the domain shift between training and application phases. Traditional unsupervised domain adaptive detection methods often struggle with this shift, leading to suboptimal outcomes. To overcome these challenges, we introduce a novel domain adaptive approach for lung nodule detection that leverages mean teacher self-training and contrastive learning. First, we propose a hierarchical contrastive learning strategy to refine nodule representations and enhance the distinction between nodules and background. Second, we introduce a nodule-level domain-invariant feature learning (NDL) module to capture domain-invariant features through adversarial learning across different domains. Additionally, we propose a new annotated dataset of X-ray images to aid in advancing lung nodule detection research. Extensive experiments conducted on multiple X-ray datasets demonstrate the efficacy of our approach in mitigating domain shift impacts.
Authors: Xihua Sheng, Chuanbo Tang, Li Li, Dong Liu, Feng Wu
Abstract: The emerging large models have achieved notable progress in the fields of natural language processing and computer vision. However, large models for neural video coding are still unexplored. In this paper, we try to explore how to build a large neural video coding model. Based on a small baseline model, we gradually scale up the model sizes of its different coding parts, including the motion encoder-decoder, motion entropy model, contextual encoder-decoder, contextual entropy model, and temporal context mining module, and analyze the influence of model sizes on video compression performance. Then, we explore to use different architectures, including CNN, mixed CNN-Transformer, and Transformer architectures, to implement the neural video coding model and analyze the influence of model architectures on video compression performance. Based on our exploration results, we design the first neural video coding model with more than 1 billion parameters -- NVC-1B. Experimental results show that our proposed large model achieves a significant video compression performance improvement over the small baseline model, and represents the state-of-the-art compression efficiency. We anticipate large models may bring up the video coding technologies to the next level.
Authors: Qiao Li, Kanlun Tan, Qiao Liu, Di Yuan, Xin Li, Yunpeng Liu
Abstract: Due to the lack of large-scale labeled Thermal InfraRed (TIR) training datasets, most existing TIR trackers are trained directly on RGB datasets. However, tracking methods trained on RGB datasets suffer a significant drop-off in TIR data due to the domain shift issue. To this end, in this work, we propose a Progressive Domain Adaptation framework for TIR Tracking (PDAT), which transfers useful knowledge learned from RGB tracking to TIR tracking. The framework makes full use of large-scale labeled RGB datasets without requiring time-consuming and labor-intensive labeling of large-scale TIR data. Specifically, we first propose an adversarial-based global domain adaptation module to reduce domain gap on the feature level coarsely. Second, we design a clustering-based subdomain adaptation method to further align the feature distributions of the RGB and TIR datasets finely. These two domain adaptation modules gradually eliminate the discrepancy between the two domains, and thus learn domain-invariant fine-grained features through progressive training. Additionally, we collect a largescale TIR dataset with over 1.48 million unlabeled TIR images for training the proposed domain adaptation framework. Experimental results on five TIR tracking benchmarks show that the proposed method gains a nearly 6% success rate, demonstrating its effectiveness.
Authors: Hao Zhu, Zhen Liu, Qi Zhang, Jingde Fu, Weibing Deng, Zhan Ma, Yanwen Guo, Xun Cao
Abstract: Implicit Neural Representation (INR), which utilizes a neural network to map coordinate inputs to corresponding attributes, is causing a revolution in the field of signal processing. However, current INR techniques suffer from the "frequency"-specified spectral bias and capacity-convergence gap, resulting in imperfect performance when representing complex signals with multiple "frequencies". We have identified that both of these two characteristics could be handled by increasing the utilization of definition domain in current activation functions, for which we propose the FINER++ framework by extending existing periodic/non-periodic activation functions to variable-periodic ones. By initializing the bias of the neural network with different ranges, sub-functions with various frequencies in the variable-periodic function are selected for activation. Consequently, the supported frequency set can be flexibly tuned, leading to improved performance in signal representation. We demonstrate the generalization and capabilities of FINER++ with different activation function backbones (Sine, Gauss. and Wavelet) and various tasks (2D image fitting, 3D signed distance field representation, 5D neural radiance fields optimization and streamable INR transmission), and we show that it improves existing INRs. Project page: {https://liuzhen0212.github.io/finerpp/}
Authors: Zhen Chen, Zongming Zhang, Wenwu Guo, Xingjian Luo, Long Bai, Jinlin Wu, Hongliang Ren, Hongbin Liu
Abstract: Surgical instrument segmentation is crucial in surgical scene understanding, thereby facilitating surgical safety. Existing algorithms directly detected all instruments of pre-defined categories in the input image, lacking the capability to segment specific instruments according to the surgeon's intention. During different stages of surgery, surgeons exhibit varying preferences and focus toward different surgical instruments. Therefore, an instrument segmentation algorithm that adheres to the surgeon's intention can minimize distractions from irrelevant instruments and assist surgeons to a great extent. The recent Segment Anything Model (SAM) reveals the capability to segment objects following prompts, but the manual annotations for prompts are impractical during the surgery. To address these limitations in operating rooms, we propose an audio-driven surgical instrument segmentation framework, named ASI-Seg, to accurately segment the required surgical instruments by parsing the audio commands of surgeons. Specifically, we propose an intention-oriented multimodal fusion to interpret the segmentation intention from audio commands and retrieve relevant instrument details to facilitate segmentation. Moreover, to guide our ASI-Seg segment of the required surgical instruments, we devise a contrastive learning prompt encoder to effectively distinguish the required instruments from the irrelevant ones. Therefore, our ASI-Seg promotes the workflow in the operating rooms, thereby providing targeted support and reducing the cognitive load on surgeons. Extensive experiments are performed to validate the ASI-Seg framework, which reveals remarkable advantages over classical state-of-the-art and medical SAMs in both semantic segmentation and intention-oriented segmentation. The source code is available at https://github.com/Zonmgin-Zhang/ASI-Seg.
Authors: Zhongling Huang, Yihan Zhuang, Zipei Zhong, Feng Xu, Gong Cheng, Junwei Han
Abstract: SAR image simulation has attracted much attention due to its great potential to supplement the scarce training data for deep learning algorithms. Consequently, evaluating the quality of the simulated SAR image is crucial for practical applications. The current literature primarily uses image quality assessment techniques for evaluation that rely on human observers' perceptions. However, because of the unique imaging mechanism of SAR, these techniques may produce evaluation results that are not entirely valid. The distribution inconsistency between real and simulated data is the main obstacle that influences the utility of simulated SAR images. To this end, we propose a novel trustworthy utility evaluation framework with a counterfactual explanation for simulated SAR images for the first time, denoted as X-Fake. It unifies a probabilistic evaluator and a causal explainer to achieve a trustworthy utility assessment. We construct the evaluator using a probabilistic Bayesian deep model to learn the posterior distribution, conditioned on real data. Quantitatively, the predicted uncertainty of simulated data can reflect the distribution discrepancy. We build the causal explainer with an introspective variational auto-encoder to generate high-resolution counterfactuals. The latent code of IntroVAE is finally optimized with evaluation indicators and prior information to generate the counterfactual explanation, thus revealing the inauthentic details of simulated data explicitly. The proposed framework is validated on four simulated SAR image datasets obtained from electromagnetic models and generative artificial intelligence approaches. The results demonstrate the proposed X-Fake framework outperforms other IQA methods in terms of utility. Furthermore, the results illustrate that the generated counterfactual explanations are trustworthy, and can further improve the data utility in applications.
Authors: Chengan He, Xin Sun, Zhixin Shu, Fujun Luan, S\"oren Pirk, Jorge Alejandro Amador Herrera, Dominik L. Michels, Tuanfeng Y. Wang, Meng Zhang, Holly Rushmeier, Yi Zhou
Abstract: We present \textsc{Perm}, a learned parametric model of human 3D hair designed to facilitate various hair-related applications. Unlike previous work that jointly models the global hair shape and local strand details, we propose to disentangle them using a PCA-based strand representation in the frequency domain, thereby allowing more precise editing and output control. Specifically, we leverage our strand representation to fit and decompose hair geometry textures into low- to high-frequency hair structures. These decomposed textures are later parameterized with different generative models, emulating common stages in the hair modeling process. We conduct extensive experiments to validate the architecture design of \textsc{Perm}, and finally deploy the trained model as a generic prior to solve task-agnostic problems, further showcasing its flexibility and superiority in tasks such as 3D hair parameterization, hairstyle interpolation, single-view hair reconstruction, and hair-conditioned image generation. Our code and data will be available at: \url{https://github.com/c-he/perm}.
Authors: Changgu Chen, Libing Yang, Xiaoyan Yang, Lianggangxu Chen, Gaoqi He, CHangbo Wang, Yang Li
Abstract: In recent years, large-scale pre-trained diffusion models have demonstrated their outstanding capabilities in image and video generation tasks. However, existing models tend to produce visual objects commonly found in the training dataset, which diverges from user input prompts. The underlying reason behind the inaccurate generated results lies in the model's difficulty in sampling from specific intervals of the initial noise distribution corresponding to the prompt. Moreover, it is challenging to directly optimize the initial distribution, given that the diffusion process involves multiple denoising steps. In this paper, we introduce a Fine-tuning Initial Noise Distribution (FIND) framework with policy optimization, which unleashes the powerful potential of pre-trained diffusion networks by directly optimizing the initial distribution to align the generated contents with user-input prompts. To this end, we first reformulate the diffusion denoising procedure as a one-step Markov decision process and employ policy optimization to directly optimize the initial distribution. In addition, a dynamic reward calibration module is proposed to ensure training stability during optimization. Furthermore, we introduce a ratio clipping algorithm to utilize historical data for network training and prevent the optimized distribution from deviating too far from the original policy to restrain excessive optimization magnitudes. Extensive experiments demonstrate the effectiveness of our method in both text-to-image and text-to-video tasks, surpassing SOTA methods in achieving consistency between prompts and the generated content. Our method achieves 10 times faster than the SOTA approach. Our homepage is available at \url{https://github.com/vpx-ecnu/FIND-website}.
Authors: Yui Lo, Yuqian Chen, Fan Zhang, Dongnan Liu, Leo Zekelman, Suheyla Cetin-Karayumak, Yogesh Rathi, Weidong Cai, Lauren J. O'Donnell
Abstract: Parcellation of white matter tractography provides anatomical features for disease prediction, anatomical tract segmentation, surgical brain mapping, and non-imaging phenotype classifications. However, parcellation does not always reach 100% accuracy due to various factors, including inter-individual anatomical variability and the quality of neuroimaging scan data. The failure to identify parcels causes a problem of missing microstructure data values, which is especially challenging for downstream tasks that analyze large brain datasets. In this work, we propose a novel deep-learning model to impute tissue microstructure: the White Matter Geometry-guided Diffusion (WMG-Diff) model. Specifically, we first propose a deep score-based guided diffusion model to impute tissue microstructure for diffusion magnetic resonance imaging (dMRI) tractography fiber clusters. Second, we propose a white matter atlas geometric relationship-guided denoising function to guide the reverse denoising process at the subject-specific level. Third, we train and evaluate our model on a large dataset with 9342 subjects. Comprehensive experiments for tissue microstructure imputation and a downstream non-imaging phenotype prediction task demonstrate that our proposed WMG-Diff outperforms state-of-the-art methods.
Authors: Buyu Liu, Kai Wang, Yansong Liu, Jun Bao, Tingting Han, Jun Yu
Abstract: This work aims to address the multi-view perspective RGB generation from text prompts given Bird-Eye-View(BEV) semantics. Unlike prior methods that neglect layout consistency, lack the ability to handle detailed text prompts, or are incapable of generalizing to unseen view points, MVPbev simultaneously generates cross-view consistent images of different perspective views with a two-stage design, allowing object-level control and novel view generation at test-time. Specifically, MVPbev firstly projects given BEV semantics to perspective view with camera parameters, empowering the model to generalize to unseen view points. Then we introduce a multi-view attention module where special initialization and de-noising processes are introduced to explicitly enforce local consistency among overlapping views w.r.t. cross-view homography. Last but not least, MVPbev further allows test-time instance-level controllability by refining a pre-trained text-to-image diffusion model. Our extensive experiments on NuScenes demonstrate that our method is capable of generating high-resolution photorealistic images from text descriptions with thousands of training samples, surpassing the state-of-the-art methods under various evaluation metrics. We further demonstrate the advances of our method in terms of generalizability and controllability with the help of novel evaluation metrics and comprehensive human analysis. Our code, data, and model can be found in \url{https://github.com/kkaiwwana/MVPbev}.
Authors: Jiarui Duan, Haoling Li, Haofei Zhang, Hao Jiang, Mengqi Xue, Li Sun, Mingli Song, Jie Song
Abstract: Attribution-based explanations are garnering increasing attention recently and have emerged as the predominant approach towards \textit{eXplanable Artificial Intelligence}~(XAI). However, the absence of consistent configurations and systematic investigations in prior literature impedes comprehensive evaluations of existing methodologies. In this work, we introduce {Meta-Rank}, an open platform for benchmarking attribution methods in the image domain. Presently, Meta-Rank assesses eight exemplary attribution methods using six renowned model architectures on four diverse datasets, employing both the \textit{Most Relevant First} (MoRF) and \textit{Least Relevant First} (LeRF) evaluation protocols. Through extensive experimentation, our benchmark reveals three insights in attribution evaluation endeavors: 1) evaluating attribution methods under disparate settings can yield divergent performance rankings; 2) although inconsistent across numerous cases, the performance rankings exhibit remarkable consistency across distinct checkpoints along the same training trajectory; 3) prior attempts at consistent evaluation fare no better than baselines when extended to more heterogeneous models and datasets. Our findings underscore the necessity for future research in this domain to conduct rigorous evaluations encompassing a broader range of models and datasets, and to reassess the assumptions underlying the empirical success of different attribution methods. Our code is publicly available at \url{https://github.com/TreeThree-R/Meta-Rank}.
Authors: Fernando Alonso-Fernandez, Kevin Hernandez-Diaz, Prayag Tiwari, Josef Bigun
Abstract: We apply pre-trained architectures, originally developed for the ImageNet Large Scale Visual Recognition Challenge, for periocular recognition. These architectures have demonstrated significant success in various computer vision tasks beyond the ones for which they were designed. This work builds on our previous study using off-the-shelf Convolutional Neural Network (CNN) and extends it to include the more recently proposed Vision Transformers (ViT). Despite being trained for generic object classification, middle-layer features from CNNs and ViTs are a suitable way to recognize individuals based on periocular images. We also demonstrate that CNNs and ViTs are highly complementary since their combination results in boosted accuracy. In addition, we show that a small portion of these pre-trained models can achieve good accuracy, resulting in thinner models with fewer parameters, suitable for resource-limited environments such as mobiles. This efficiency improves if traditional handcrafted features are added as well.
Authors: Nitzan Bitton-Guetta, Aviv Slobodkin, Aviya Maimon, Eliya Habba, Royi Rassin, Yonatan Bitton, Idan Szpektor, Amir Globerson, Yuval Elovici
Abstract: Imagine observing someone scratching their arm; to understand why, additional context would be necessary. However, spotting a mosquito nearby would immediately offer a likely explanation for the person's discomfort, thereby alleviating the need for further information. This example illustrates how subtle visual cues can challenge our cognitive skills and demonstrates the complexity of interpreting visual scenarios. To study these skills, we present Visual Riddles, a benchmark aimed to test vision and language models on visual riddles requiring commonsense and world knowledge. The benchmark comprises 400 visual riddles, each featuring a unique image created by a variety of text-to-image models, question, ground-truth answer, textual hint, and attribution. Human evaluation reveals that existing models lag significantly behind human performance, which is at 82\% accuracy, with Gemini-Pro-1.5 leading with 40\% accuracy. Our benchmark comes with automatic evaluation tasks to make assessment scalable. These findings underscore the potential of Visual Riddles as a valuable resource for enhancing vision and language models' capabilities in interpreting complex visual scenarios.
Authors: Chenhao Wang, Xiaopeng Hong, Zhiheng Ma, Yupeng Wei, Yabin Wang, Xiaopeng Fan
Abstract: Multi-modal crowd counting is a crucial task that uses multi-modal cues to estimate the number of people in crowded scenes. To overcome the gap between different modalities, we propose a modal emulation-based two-pass multi-modal crowd-counting framework that enables efficient modal emulation, alignment, and fusion. The framework consists of two key components: a \emph{multi-modal inference} pass and a \emph{cross-modal emulation} pass. The former utilizes a hybrid cross-modal attention module to extract global and local information and achieve efficient multi-modal fusion. The latter uses attention prompting to coordinate different modalities and enhance multi-modal alignment. We also introduce a modality alignment module that uses an efficient modal consistency loss to align the outputs of the two passes and bridge the semantic gap between modalities. Extensive experiments on both RGB-Thermal and RGB-Depth counting datasets demonstrate its superior performance compared to previous methods. Code available at https://github.com/Mr-Monday/Multi-modal-Crowd-Counting-via-Modal-Emulation.
URLs: https://github.com/Mr-Monday/Multi-modal-Crowd-Counting-via-Modal-Emulation.
Authors: Yihao Wang, Lizhi Chen, Zhong Qian, Peifeng Li
Abstract: News media, especially video news media, have penetrated into every aspect of daily life, which also brings the risk of fake news. Therefore, multimodal fake news detection has recently received more attention. However, the number of fake news detection data sets for video modal is small, and these data sets are composed of unofficial videos uploaded by users, so there is too much useless data. To solve this problem, we present in this paper a dataset named Official-NV, which consists of officially published news videos on Xinhua. We crawled videos on Xinhua, and then extended the data set using LLM generation and manual modification. In addition, we benchmarked the data set presented in this paper using a baseline model to demonstrate the advantage of Official-NV in multimodal fake news detection.
Authors: Zhengcen Li, Xinle Chang, Yueran Li, Jingyong Su
Abstract: Group Activity Recognition aims to understand collective activities from videos. Existing solutions primarily rely on the RGB modality, which encounters challenges such as background variations, occlusions, motion blurs, and significant computational overhead. Meanwhile, current keypoint-based methods offer a lightweight and informative representation of human motions but necessitate accurate individual annotations and specialized interaction reasoning modules. To address these limitations, we design a panoramic graph that incorporates multi-person skeletons and objects to encapsulate group activity, offering an effective alternative to RGB video. This panoramic graph enables Graph Convolutional Network (GCN) to unify intra-person, inter-person, and person-object interactive modeling through spatial-temporal graph convolutions. In practice, we develop a novel pipeline that extracts skeleton coordinates using pose estimation and tracking algorithms and employ Multi-person Panoramic GCN (MP-GCN) to predict group activities. Extensive experiments on Volleyball and NBA datasets demonstrate that the MP-GCN achieves state-of-the-art performance in both accuracy and efficiency. Notably, our method outperforms RGB-based approaches by using only estimated 2D keypoints as input. Code is available at https://github.com/mgiant/MP-GCN
Authors: Jingjing Wu, Zhengyao Fang, Pengyuan Lyu, Chengquan Zhang, Fanglin Chen, Guangming Lu, Wenjie Pei
Abstract: Transcription-only Supervised Text Spotting aims to learn text spotters relying only on transcriptions but no text boundaries for supervision, thus eliminating expensive boundary annotation. The crux of this task lies in locating each transcription in scene text images without location annotations. In this work, we formulate this challenging problem as a Weakly Supervised Cross-modality Contrastive Learning problem, and design a simple yet effective model dubbed WeCromCL that is able to detect each transcription in a scene image in a weakly supervised manner. Unlike typical methods for cross-modality contrastive learning that focus on modeling the holistic semantic correlation between an entire image and a text description, our WeCromCL conducts atomistic contrastive learning to model the character-wise appearance consistency between a text transcription and its correlated region in a scene image to detect an anchor point for the transcription in a weakly supervised manner. The detected anchor points by WeCromCL are further used as pseudo location labels to guide the learning of text spotting. Extensive experiments on four challenging benchmarks demonstrate the superior performance of our model over other methods. Code will be released.
Authors: Honglin Li, Yusuan Sun, Chenglu Zhu, Yunlong Zhang, Shichuan Zhang, Zhongyi Shui, Pingyi Chen, Jingxiong Li, Sunyi Zheng, Can Cui, Lin Yang
Abstract: Cervical Cancer continues to be the leading gynecological malignancy, posing a persistent threat to women's health on a global scale. Early screening via cytology Whole Slide Image (WSI) diagnosis is critical to prevent this Cancer progression and improve survival rate, but pathologist's single test suffers inevitable false negative due to the immense number of cells that need to be reviewed within a WSI. Though computer-aided automated diagnostic models can serve as strong complement for pathologists, their effectiveness is hampered by the paucity of extensive and detailed annotations, coupled with the limited interpretability and robustness. These factors significantly hinder their practical applicability and reliability in clinical settings. To tackle these challenges, we develop an AI approach, which is a Scalable Technology for Robust and Interpretable Diagnosis built on Extensive data (STRIDE) of cervical cytology. STRIDE addresses the bottleneck of limited annotations by integrating patient-level labels with a small portion of cell-level labels through an end-to-end training strategy, facilitating scalable learning across extensive datasets. To further improve the robustness to real-world domain shifts of cytology slide-making and imaging, STRIDE employs color adversarial samples training that mimic staining and imaging variations. Lastly, to achieve pathologist-level interpretability for the trustworthiness in clinical settings, STRIDE can generate explanatory textual descriptions that simulates pathologists' diagnostic processes by cell image feature and textual description alignment. Conducting extensive experiments and evaluations in 183 medical centers with a dataset of 341,889 WSIs and 0.1 billion cells from cervical cytology patients, STRIDE has demonstrated a remarkable superiority over previous state-of-the-art techniques.
Authors: Yunfeng Fan, Wenchao Xu, Haozhao Wang, Junhong Liu, Song Guo
Abstract: Recently, Multimodal Learning (MML) has gained significant interest as it compensates for single-modality limitations through comprehensive complementary information within multimodal data. However, traditional MML methods generally use the joint learning framework with a uniform learning objective that can lead to the modality competition issue, where feedback predominantly comes from certain modalities, limiting the full potential of others. In response to this challenge, this paper introduces DI-MML, a novel detached MML framework designed to learn complementary information across modalities under the premise of avoiding modality competition. Specifically, DI-MML addresses competition by separately training each modality encoder with isolated learning objectives. It further encourages cross-modal interaction via a shared classifier that defines a common feature space and employing a dimension-decoupled unidirectional contrastive (DUC) loss to facilitate modality-level knowledge transfer. Additionally, to account for varying reliability in sample pairs, we devise a certainty-aware logit weighting strategy to effectively leverage complementary information at the instance level during inference. Extensive experiments conducted on audio-visual, flow-image, and front-rear view datasets show the superior performance of our proposed method. The code is released at https://github.com/fanyunfeng-bit/DI-MML.
Authors: Tz-Ying Wu, Kyle Min, Subarna Tripathi, Nuno Vasconcelos
Abstract: Video understanding typically requires fine-tuning the large backbone when adapting to new domains. In this paper, we leverage the egocentric video foundation models (Ego-VFMs) based on video-language pre-training and propose a parameter-efficient adaptation for egocentric video tasks, namely Ego-VPA. It employs a local sparse approximation for each video frame/text feature using the basis prompts, and the selected basis prompts are used to synthesize video/text prompts. Since the basis prompts are shared across frames and modalities, it models context fusion and cross-modal transfer in an efficient fashion. Experiments show that Ego-VPA excels in lightweight adaptation (with only 0.84% learnable parameters), largely improving over baselines and reaching the performance of full fine-tuning.
Authors: Hanjun Luo, Ziye Deng, Haoyu Huang, Xuecheng Liu, Ruizhe Chen, Zuozhu Liu
Abstract: With the rapid development of Text-to-Image models, biases in human image generation against demographic groups social attract more and more concerns. Existing methods are designed based on certain models with fixed prompts, unable to accommodate the trend of high-speed updating of Text-to-Image (T2I) models and variable prompts in practical scenes. Additionally, they fail to consider the possibility of hallucinations, leading to deviations between expected and actual results. To address this issue, we introduce VersusDebias, a novel and universal debiasing framework for biases in T2I models, consisting of one generative adversarial mechanism (GAM) and one debiasing generation mechanism using a small language model (SLM). The self-adaptive GAM generates specialized attribute arrays for each prompts for diminishing the influence of hallucinations from T2I models. The SLM uses prompt engineering to generate debiased prompts for the T2I model, providing zero-shot debiasing ability and custom optimization for different models. Extensive experiments demonstrate VersusDebias's capability to rectify biases on arbitrary models across multiple protected attributes simultaneously, including gender, race, and age. Furthermore, VersusDebias outperforms existing methods in both zero-shot and few-shot situations, illustrating its extraordinary utility. Our work is openly accessible to the research community to ensure the reproducibility.
Authors: Shuang Wu, Songlin Tang, Guangming Lu, Jianzhuang Liu, Wenjie Pei
Abstract: Typical inverse rendering methods focus on learning implicit neural scene representations by modeling the geometry, materials and illumination separately, which entails significant computations for optimization. In this work we design a Unified Voxelization framework for explicit learning of scene representations, dubbed UniVoxel, which allows for efficient modeling of the geometry, materials and illumination jointly, thereby accelerating the inverse rendering significantly. To be specific, we propose to encode a scene into a latent volumetric representation, based on which the geometry, materials and illumination can be readily learned via lightweight neural networks in a unified manner. Particularly, an essential design of UniVoxel is that we leverage local Spherical Gaussians to represent the incident light radiance, which enables the seamless integration of modeling illumination into the unified voxelization framework. Such novel design enables our UniVoxel to model the joint effects of direct lighting, indirect lighting and light visibility efficiently without expensive multi-bounce ray tracing. Extensive experiments on multiple benchmarks covering diverse scenes demonstrate that UniVoxel boosts the optimization efficiency significantly compared to other methods, reducing the per-scene training time from hours to 18 minutes, while achieving favorable reconstruction quality. Code is available at https://github.com/freemantom/UniVoxel.
Authors: Biao Wu, Yutong Xie, Zeyu Zhang, Minh Hieu Phan, Qi Chen, Ling Chen, Qi Wu
Abstract: Vision-and-language pretraining (VLP) in the medical field utilizes contrastive learning on image-text pairs to achieve effective transfer across tasks. Yet, current VLP approaches with the masked modelling strategy face two challenges when applied to the medical domain. First, current models struggle to accurately reconstruct key pathological features due to the scarcity of medical data. Second, most methods only adopt either paired image-text or image-only data, failing to exploit the combination of both paired and unpaired data. To this end, this paper proposes a XLIP (Masked modelling for medical Language-Image Pre-training) framework to enhance pathological learning and feature learning via unpaired data. First, we introduce the attention-masked image modelling (AttMIM) and entity-driven masked language modelling module (EntMLM), which learns to reconstruct pathological visual and textual tokens via multi-modal feature interaction, thus improving medical-enhanced features. The AttMIM module masks a portion of the image features that are highly responsive to textual features. This allows XLIP to improve the reconstruction of highly similar image data in medicine efficiency. Second, our XLIP capitalizes unpaired data to enhance multimodal learning by introducing disease-kind prompts. The experimental results show that XLIP achieves SOTA for zero-shot and fine-tuning classification performance on five datasets. Our code will be available at https://github.com/White65534/XLIP
Authors: Yushi Huang, Ruihao Gong, Xianglong Liu, Jing Liu, Yuhang Li, Jiwen Lu, Dacheng Tao
Abstract: The Diffusion models, widely used for image generation, face significant challenges related to their broad applicability due to prolonged inference times and high memory demands. Efficient Post-Training Quantization (PTQ) is crucial to address these issues in traditional models. Unlike those models, diffusion models critically rely on the time-step $t$ for effective multi-round denoising. Typically, $t$ from the finite set $\{1, \ldots, T\}$ is encoded into a hypersensitive temporal feature by several modules, entirely independent of the sampling data. However, existing PTQ methods do not optimize these modules individually. Instead, they employ unsuitable reconstruction objectives and complex calibration methods, leading to significant disturbances in the temporal feature and denoising trajectory. To address these challenges, we introduce a novel quantization framework: 1)~TIB-based Maintenance: Based on our innovative Temporal Information Block~(TIB) definition, Temporal Information-aware Reconstruction~(TIAR) and Finite Set Calibration~(FSC) are developed to efficiently align full precision temporal features. 2)~Cache-based Maintenance: Instead of indirect and complex optimization for the related modules, pre-computing and caching quantized counterparts of temporal features are developed to minimize errors. 3)~Disturbance-aware Selection: Employ temporal feature errors to guide a fine-grained selection for superior maintenance. This framework preserves most of the temporal information and ensures high-quality end-to-end generation. Extensive testing on various datasets and diffusion models confirms our superior results. Notably, our approach closely matches the performance of the full-precision model under 4-bit quantization. Furthermore, the quantized SD-XL model achieves hardware acceleration of 2.20$\times$ on CPU and 5.76$\times$ on GPU demonstrating its efficiency.
Authors: Zhenyu Tang, Junwu Zhang, Xinhua Cheng, Wangbo Yu, Chaoran Feng, Yatian Pang, Bin Lin, Li Yuan
Abstract: Recent 3D large reconstruction models typically employ a two-stage process, including first generate multi-view images by a multi-view diffusion model, and then utilize a feed-forward model to reconstruct images to 3D content.However, multi-view diffusion models often produce low-quality and inconsistent images, adversely affecting the quality of the final 3D reconstruction. To address this issue, we propose a unified 3D generation framework called Cycle3D, which cyclically utilizes a 2D diffusion-based generation module and a feed-forward 3D reconstruction module during the multi-step diffusion process. Concretely, 2D diffusion model is applied for generating high-quality texture, and the reconstruction model guarantees multi-view consistency.Moreover, 2D diffusion model can further control the generated content and inject reference-view information for unseen views, thereby enhancing the diversity and texture consistency of 3D generation during the denoising process. Extensive experiments demonstrate the superior ability of our method to create 3D content with high-quality and consistency compared with state-of-the-art baselines.
Authors: Vikash Kumar, Himanshu Patil, Rohit Lal, Anirban Chakraborty
Abstract: In this work, we explore the usage of the Frequency Transformation for reducing the domain shift between the source and target domain (e.g., synthetic image and real image respectively) towards solving the Domain Adaptation task. Most of the Unsupervised Domain Adaptation (UDA) algorithms focus on reducing the global domain shift between labelled source and unlabelled target domains by matching the marginal distributions under a small domain gap assumption. UDA performance degrades for the cases where the domain gap between source and target distribution is large. In order to bring the source and the target domains closer, we propose a novel approach based on traditional image processing technique Class Aware Frequency Transformation (CAFT) that utilizes pseudo label based class consistent low-frequency swapping for improving the overall performance of the existing UDA algorithms. The proposed approach, when compared with the state-of-the-art deep learning based methods, is computationally more efficient and can easily be plugged into any existing UDA algorithm to improve its performance. Additionally, we introduce a novel approach based on absolute difference of top-2 class prediction probabilities (ADT2P) for filtering target pseudo labels into clean and noisy sets. Samples with clean pseudo labels can be used to improve the performance of unsupervised learning algorithms. We name the overall framework as CAFT++. We evaluate the same on the top of different UDA algorithms across many public domain adaptation datasets. Our extensive experiments indicate that CAFT++ is able to achieve significant performance gains across all the popular benchmarks.
Authors: Vincenzo De Rosa, Fabrizio Guillaro, Giovanni Poggi, Davide Cozzolino, Luisa Verdoliva
Abstract: In recent years, many forensic detectors have been proposed to detect AI-generated images and prevent their use for malicious purposes. Convolutional neural networks (CNNs) have long been the dominant architecture in this field and have been the subject of intense study. However, recently proposed Transformer-based detectors have been shown to match or even outperform CNN-based detectors, especially in terms of generalization. In this paper, we study the adversarial robustness of AI-generated image detectors, focusing on Contrastive Language-Image Pretraining (CLIP)-based methods that rely on Visual Transformer backbones and comparing their performance with CNN-based methods. We study the robustness to different adversarial attacks under a variety of conditions and analyze both numerical results and frequency-domain patterns. CLIP-based detectors are found to be vulnerable to white-box attacks just like CNN-based detectors. However, attacks do not easily transfer between CNN-based and CLIP-based methods. This is also confirmed by the different distribution of the adversarial noise patterns in the frequency domain. Overall, this analysis provides new insights into the properties of forensic detectors that can help to develop more effective strategies.
Authors: Jifeng Wang, Kaouther Messaoud, Yuejiang Liu, Juergen Gall, Alexandre Alahi
Abstract: Recent progress in motion forecasting has been substantially driven by self-supervised pre-training. However, adapting pre-trained models for specific downstream tasks, especially motion prediction, through extensive fine-tuning is often inefficient. This inefficiency arises because motion prediction closely aligns with the masked pre-training tasks, and traditional full fine-tuning methods fail to fully leverage this alignment. To address this, we introduce Forecast-PEFT, a fine-tuning strategy that freezes the majority of the model's parameters, focusing adjustments on newly introduced prompts and adapters. This approach not only preserves the pre-learned representations but also significantly reduces the number of parameters that need retraining, thereby enhancing efficiency. This tailored strategy, supplemented by our method's capability to efficiently adapt to different datasets, enhances model efficiency and ensures robust performance across datasets without the need for extensive retraining. Our experiments show that Forecast-PEFT outperforms traditional full fine-tuning methods in motion prediction tasks, achieving higher accuracy with only 17% of the trainable parameters typically required. Moreover, our comprehensive adaptation, Forecast-FT, further improves prediction performance, evidencing up to a 9.6% enhancement over conventional baseline methods. Code will be available at https://github.com/csjfwang/Forecast-PEFT.
Authors: ShahRukh Athar, Shunsuke Saito, Zhengyu Yang, Stanislav Pidhorsky, Chen Cao
Abstract: Creating photorealistic avatars for individuals traditionally involves extensive capture sessions with complex and expensive studio devices like the LightStage system. While recent strides in neural representations have enabled the generation of photorealistic and animatable 3D avatars from quick phone scans, they have the capture-time lighting baked-in, lack facial details and have missing regions in areas such as the back of the ears. Thus, they lag in quality compared to studio-captured avatars. In this paper, we propose a method that bridges this gap by generating studio-like illuminated texture maps from short, monocular phone captures. We do this by parameterizing the phone texture maps using the $W^+$ space of a StyleGAN2, enabling near-perfect reconstruction. Then, we finetune a StyleGAN2 by sampling in the $W^+$ parameterized space using a very small set of studio-captured textures as an adversarial training signal. To further enhance the realism and accuracy of facial details, we super-resolve the output of the StyleGAN2 using carefully designed diffusion model that is guided by image gradients of the phone-captured texture map. Once trained, our method excels at producing studio-like facial texture maps from casual monocular smartphone videos. Demonstrating its capabilities, we showcase the generation of photorealistic, uniformly lit, complete avatars from monocular phone captures. The project page can be found at http://shahrukhathar.github.io/2024/07/22/Bridging.html
URLs: http://shahrukhathar.github.io/2024/07/22/Bridging.html
Authors: Sounak Mondal, Seoyoung Ahn, Zhibo Yang, Niranjan Balasubramanian, Dimitris Samaras, Gregory Zelinsky, Minh Hoai
Abstract: For computer systems to effectively interact with humans using spoken language, they need to understand how the words being generated affect the users' moment-by-moment attention. Our study focuses on the incremental prediction of attention as a person is seeing an image and hearing a referring expression defining the object in the scene that should be fixated by gaze. To predict the gaze scanpaths in this incremental object referral task, we developed the Attention in Referral Transformer model or ART, which predicts the human fixations spurred by each word in a referring expression. ART uses a multimodal transformer encoder to jointly learn gaze behavior and its underlying grounding tasks, and an autoregressive transformer decoder to predict, for each word, a variable number of fixations based on fixation history. To train ART, we created RefCOCO-Gaze, a large-scale dataset of 19,738 human gaze scanpaths, corresponding to 2,094 unique image-expression pairs, from 220 participants performing our referral task. In our quantitative and qualitative analyses, ART not only outperforms existing methods in scanpath prediction, but also appears to capture several human attention patterns, such as waiting, scanning, and verification.
Authors: Yang Wu, Kaihua Zhang, Jianjun Qian, Jin Xie, Jian Yang
Abstract: The complex traffic environment and various weather conditions make the collection of LiDAR data expensive and challenging. Achieving high-quality and controllable LiDAR data generation is urgently needed, controlling with text is a common practice, but there is little research in this field. To this end, we propose Text2LiDAR, the first efficient, diverse, and text-controllable LiDAR data generation model. Specifically, we design an equirectangular transformer architecture, utilizing the designed equirectangular attention to capture LiDAR features in a manner with data characteristics. Then, we design a control-signal embedding injector to efficiently integrate control signals through the global-to-focused attention mechanism. Additionally, we devise a frequency modulator to assist the model in recovering high-frequency details, ensuring the clarity of the generated point cloud. To foster development in the field and optimize text-controlled generation performance, we construct nuLiDARtext which offers diverse text descriptors for 34,149 LiDAR point clouds from 850 scenes. Experiments on uncontrolled and text-controlled generation in various forms on KITTI-360 and nuScenes datasets demonstrate the superiority of our approach.
Authors: Yuheng Shi, Tong Zhang, Xiaojie Guo
Abstract: Compared with still image object detection, video object detection (VOD) needs to particularly concern the high across-frame variation in object appearance, and the diverse deterioration in some frames. In principle, the detection in a certain frame of a video can benefit from information in other frames. Thus, how to effectively aggregate features across different frames is key to the target problem. Most of contemporary aggregation methods are tailored for two-stage detectors, suffering from high computational costs due to the dual-stage nature. On the other hand, although one-stage detectors have made continuous progress in handling static images, their applicability to VOD lacks sufficient exploration. To tackle the above issues, this study invents a very simple yet potent strategy of feature selection and aggregation, gaining significant accuracy at marginal computational expense. Concretely, for cutting the massive computation and memory consumption from the dense prediction characteristic of one-stage object detectors, we first condense candidate features from dense prediction maps. Then, the relationship between a target frame and its reference frames is evaluated to guide the aggregation. Comprehensive experiments and ablation studies are conducted to validate the efficacy of our design, and showcase its advantage over other cutting-edge VOD methods in both effectiveness and efficiency. Notably, our model reaches \emph{a new record performance, i.e., 92.9\% AP50 at over 30 FPS on the ImageNet VID dataset on a single 3090 GPU}, making it a compelling option for large-scale or real-time applications. The implementation is simple, and accessible at \url{https://github.com/YuHengsss/YOLOV}.
Authors: Chia-Hao Kao, Cheng Chien, Yu-Jen Tseng, Yi-Hsin Chen, Alessandro Gnutti, Shao-Yuan Lo, Wen-Hsiao Peng, Riccardo Leonardi
Abstract: This paper presents the first-ever study of adapting compressed image latents to suit the needs of downstream vision tasks that adopt Multimodal Large Language Models (MLLMs). MLLMs have extended the success of large language models to modalities (e.g. images) beyond text, but their billion scale hinders deployment on resource-constrained end devices. While cloud-hosted MLLMs could be available, transmitting raw, uncompressed images captured by end devices to the cloud requires an efficient image compression system. To address this, we focus on emerging neural image compression and propose a novel framework with a lightweight transform-neck and a surrogate loss to adapt compressed image latents for MLLM-based vision tasks. The proposed framework is generic and applicable to multiple application scenarios, where the neural image codec can be (1) pre-trained for human perception without updating, (2) fully updated for joint human and machine perception, or (3) fully updated for only machine perception. The transform-neck trained with the surrogate loss is universal, for it can serve various downstream vision tasks enabled by a variety of MLLMs that share the same visual encoder. Our framework has the striking feature of excluding the downstream MLLMs from training the transform-neck, and potentially the neural image codec as well. This stands out from most existing coding for machine approaches that involve downstream networks in training and thus could be impractical when the networks are MLLMs. Extensive experiments on different neural image codecs and various MLLM-based vision tasks show that our method achieves great rate-accuracy performance with much less complexity, demonstrating its effectiveness.
Authors: Remi Chierchia, Leo Lebrat, David Ahmedt-Aristizabal, Olivier Salvado, Clinton Fookes, Rodrigo Santa Cruz
Abstract: Managing chronic wounds is a global challenge that can be alleviated by the adoption of automatic systems for clinical wound assessment from consumer-grade videos. While 2D image analysis approaches are insufficient for handling the 3D features of wounds, existing approaches utilizing 3D reconstruction methods have not been thoroughly evaluated. To address this gap, this paper presents a comprehensive study on 3D wound reconstruction from consumer-grade videos. Specifically, we introduce the SALVE dataset, comprising video recordings of realistic wound phantoms captured with different cameras. Using this dataset, we assess the accuracy and precision of state-of-the-art methods for 3D reconstruction, ranging from traditional photogrammetry pipelines to advanced neural rendering approaches. In our experiments, we observe that photogrammetry approaches do not provide smooth surfaces suitable for precise clinical measurements of wounds. Neural rendering approaches show promise in addressing this issue, advancing the use of this technology in wound care practices.
Authors: Praveen Ravirathinam, Ankush Khandelwal, Rahul Ghosh, Vipin Kumar
Abstract: In recent years, there is increased interest in foundation models for geoscience due to vast amount of earth observing satellite imagery. Existing remote sensing foundation models make use of the various sources of spectral imagery to create large models pretrained on masked reconstruction task. The embeddings from these foundation models are then used for various downstream remote sensing applications. In this paper we propose a foundational modeling framework for remote sensing geoscience applications, that goes beyond these traditional single modality masked autoencoder family of foundation models. This framework leverages the knowledge guided principles that the spectral imagery captures the impact of the physical drivers on the environmental system, and that the relationship between them is governed by the characteristics of the system. Specifically, our method, called MultiModal Variable Step Forecasting (MM-VSF), uses mutlimodal data (spectral imagery and weather) as its input and a variable step forecasting task as its pretraining objective. In our evaluation we show forecasting of satellite imagery using weather can be used as an effective pretraining task for foundation models. We further show the effectiveness of the embeddings from MM-VSF on the downstream task of pixel wise crop mapping, when compared with a model trained in the traditional setting of single modality input and masked reconstruction based pretraining.
Authors: Mingyu Zhang, Jiting Cai, Mingyu Liu, Yue Xu, Cewu Lu, Yong-Lu Li
Abstract: Visual reasoning, as a prominent research area, plays a crucial role in AI by facilitating concept formation and interaction with the world. However, current works are usually carried out separately on small datasets thus lacking generalization ability. Through rigorous evaluation of diverse benchmarks, we demonstrate the shortcomings of existing ad-hoc methods in achieving cross-domain reasoning and their tendency to data bias fitting. In this paper, we revisit visual reasoning with a two-stage perspective: (1) symbolization and (2) logical reasoning given symbols or their representations. We find that the reasoning stage is better at generalization than symbolization. Thus, it is more efficient to implement symbolization via separated encoders for different data domains while using a shared reasoner. Given our findings, we establish design principles for visual reasoning frameworks following the separated symbolization and shared reasoning. The proposed two-stage framework achieves impressive generalization ability on various visual reasoning tasks, including puzzles, physical prediction, and visual question answering (VQA), encompassing both 2D and 3D modalities. We believe our insights will pave the way for generalizable visual reasoning.
Authors: Fangming Cui, Xun Yang, Chao Wu, Liang Xiao, Xinmei Tian
Abstract: Prompt learning represents a promising method for adapting pre-trained vision-language models (VLMs) to various downstream tasks by learning a set of text embeddings. One challenge inherent to these methods is the poor generalization performance due to the invalidity of the learned text embeddings for unseen tasks. A straightforward approach to bridge this gap is to freeze the text embeddings in prompts, which results in a lack of capacity to adapt VLMs for downstream tasks. To address this dilemma, we propose a paradigm called EnPrompt with a novel External Layer (EnLa). Specifically, we propose a textual external layer and learnable visual embeddings for adapting VLMs to downstream tasks. The learnable external layer is built upon valid embeddings of pre-trained CLIP. This design considers the balance of learning capabilities between the two branches. To align the textual and visual features, we propose a novel two-pronged approach: i) we introduce the optimal transport as the discrepancy metric to align the vision and text modalities, and ii) we introduce a novel strengthening feature to enhance the interaction between these two modalities. Four representative experiments (i.e., base-to-novel generalization, few-shot learning, cross-dataset generalization, domain shifts generalization) across 15 datasets demonstrate that our method outperforms the existing prompt learning method.
Authors: Wulian Yun, Mengshi Qi, Fei Peng, Huadong Ma
Abstract: Existing action quality assessment (AQA) methods often require a large number of label annotations for fully supervised learning, which are laborious and expensive. In practice, the labeled data are difficult to obtain because the AQA annotation process requires domain-specific expertise. In this paper, we propose a novel semi-supervised method, which can be utilized for better assessment of the AQA task by exploiting a large amount of unlabeled data and a small portion of labeled data. Differing from the traditional teacher-student network, we propose a teacher-reference-student architecture to learn both unlabeled and labeled data, where the teacher network and the reference network are used to generate pseudo-labels for unlabeled data to supervise the student network. Specifically, the teacher predicts pseudo-labels by capturing high-level features of unlabeled data. The reference network provides adequate supervision of the student network by referring to additional action information. Moreover, we introduce confidence memory to improve the reliability of pseudo-labels by storing the most accurate ever output of the teacher network and reference network. To validate our method, we conduct extensive experiments on three AQA benchmark datasets. Experimental results show that our method achieves significant improvements and outperforms existing semi-supervised AQA methods.
Authors: Hongyan Zhu, Shuai Qin, Min Su, Chengzhi Lin, Anjie Li, Junfeng Gao
Abstract: Large models can play important roles in many domains. Agriculture is another key factor affecting the lives of people around the world. It provides food, fabric, and coal for humanity. However, facing many challenges such as pests and diseases, soil degradation, global warming, and food security, how to steadily increase the yield in the agricultural sector is a problem that humans still need to solve. Large models can help farmers improve production efficiency and harvest by detecting a series of agricultural production tasks such as pests and diseases, soil quality, and seed quality. It can also help farmers make wise decisions through a variety of information, such as images, text, etc. Herein, we delve into the potential applications of large models in agriculture, from large language model (LLM) and large vision model (LVM) to large vision-language models (LVLM). After gaining a deeper understanding of multimodal large language models (MLLM), it can be recognized that problems such as agricultural image processing, agricultural question answering systems, and agricultural machine automation can all be solved by large models. Large models have great potential in the field of agriculture. We outline the current applications of agricultural large models, and aims to emphasize the importance of large models in the domain of agriculture. In the end, we envisage a future in which famers use MLLM to accomplish many tasks in agriculture, which can greatly improve agricultural production efficiency and yield.
Authors: Isaac Osei Agyemang, Jianwen Chen, Liaoyuan Zeng, Isaac Adjei-Mensah, Daniel Acheampong, Gordon Owusu Boateng, Adu Asare Baffour
Abstract: Image environments and noisy labels hinder deep learning-based inference models in structural damage detection. Post-detection, there is the challenge of reliance on manual assessments of detected damages. As a result, Guided-DetNet, characterized by Generative Attention Module (GAM), Hierarchical Elimination Algorithm (HEA), and Volumetric Contour Visual Assessment (VCVA), is proposed to mitigate complex image environments, noisy labeling, and post-detection manual assessment of structural damages. GAM leverages cross-horizontal and cross-vertical patch merging and cross foreground-background feature fusion to generate varied features to mitigate complex image environments. HEA addresses noisy labeling using hierarchical relationships among classes to refine instances given an image by eliminating unlikely class categories. VCVA assesses the severity of detected damages via volumetric representation and quantification leveraging the Dirac delta distribution. A comprehensive quantitative study, two robustness tests, and an application scenario based on the PEER Hub Image-Net dataset substantiate Guided-DetNet's promising performances. Guided-DetNet outperformed the best-compared models in a triple classification task by a difference of not less than 3% and not less than 2% in a dual detection task under varying metrics.
Authors: Zewen Du, Zhenjiang Hu, Guiyu Zhao, Ying Jin, Hongbin Ma
Abstract: Object detection in aerial images has always been a challenging task due to the generally small size of the objects. Most current detectors prioritize novel detection frameworks, often overlooking research on fundamental components such as feature pyramid networks. In this paper, we introduce the Cross-Layer Feature Pyramid Transformer (CFPT), a novel upsampler-free feature pyramid network designed specifically for small object detection in aerial images. CFPT incorporates two meticulously designed attention blocks with linear computational complexity: the Cross-Layer Channel-Wise Attention (CCA) and the Cross-Layer Spatial-Wise Attention (CSA). CCA achieves cross-layer interaction by dividing channel-wise token groups to perceive cross-layer global information along the spatial dimension, while CSA completes cross-layer interaction by dividing spatial-wise token groups to perceive cross-layer global information along the channel dimension. By integrating these modules, CFPT enables cross-layer interaction in one step, thereby avoiding the semantic gap and information loss associated with element-wise summation and layer-by-layer transmission. Furthermore, CFPT incorporates global contextual information, which enhances detection performance for small objects. To further enhance location awareness during cross-layer interaction, we propose the Cross-Layer Consistent Relative Positional Encoding (CCPE) based on inter-layer mutual receptive fields. We evaluate the effectiveness of CFPT on two challenging object detection datasets in aerial images, namely VisDrone2019-DET and TinyPerson. Extensive experiments demonstrate the effectiveness of CFPT, which outperforms state-of-the-art feature pyramid networks while incurring lower computational costs. The code will be released at https://github.com/duzw9311/CFPT.
Authors: Jinsung Lee, Taeoh Kim, Inwoong Lee, Minho Shim, Dongyoon Wee, Minsu Cho, Suha Kwak
Abstract: Video action detection (VAD) aims to detect actors and classify their actions in a video. We figure that VAD suffers more from classification rather than localization of actors. Hence, we analyze how prevailing methods form features for classification and find that they prioritize actor regions, yet often overlooking the essential contextual information necessary for accurate classification. Accordingly, we propose to reduce the bias toward actor and encourage paying attention to the context that is relevant to each action class. By assigning a class-dedicated query to each action class, our model can dynamically determine where to focus for effective classification. The proposed model demonstrates superior performance on three challenging benchmarks with significantly fewer parameters and less computation.
Authors: Ezequiel Perez-Zarate, Oscar Ramos-Soto, Diego Oliva, Marco Perez-Cisneros
Abstract: Low-light image enhancement is an important task in computer vision, essential for improving the visibility and quality of images captured in non-optimal lighting conditions. Inadequate illumination can lead to significant information loss and poor image quality, impacting various applications such as surveillance. photography, or even autonomous driving. In this regard, automated methods have been developed to automatically adjust illumination in the image for a better visual perception. Current enhancement techniques often use specific datasets to enhance low-light images, but still present challenges when adapting to diverse real-world conditions, where illumination degradation may be localized to specific regions. To address this challenge, the Adaptive Light Enhancement Network (ALEN) is introduced, whose main approach is the use of a classification mechanism to determine whether local or global illumination enhancement is required. Subsequently, estimator networks adjust illumination based on this classification and simultaneously enhance color fidelity. ALEN integrates the Light Classification Network (LCNet) for illuminance categorization, complemented by the Single-Channel Network (SCNet), and Multi-Channel Network (MCNet) for precise estimation of illumination and color, respectively. Extensive experiments on publicly available datasets for low-light conditions were carried out to underscore ALEN's robust generalization capabilities, demonstrating superior performance in both quantitative metrics and qualitative assessments when compared to recent state-of-the-art methods. The ALEN not only enhances image quality in terms of visual perception but also represents an advancement in high-level vision tasks, such as semantic segmentation, as presented in this work. The code of this method is available at https://github.com/xingyumex/ALEN.
Authors: Muhammad Abdullah Jamal, Omid Mohareri
Abstract: Surgical scene understanding is a key technical component for enabling intelligent and context aware systems that can transform various aspects of surgical interventions. In this work, we focus on the semantic segmentation task, propose a simple yet effective multi-modal (RGB and depth) training framework called SurgDepth, and show state-of-the-art (SOTA) results on all publicly available datasets applicable for this task. Unlike previous approaches, which either fine-tune SOTA segmentation models trained on natural images, or encode RGB or RGB-D information using RGB only pre-trained backbones, SurgDepth, which is built on top of Vision Transformers (ViTs), is designed to encode both RGB and depth information through a simple fusion mechanism. We conduct extensive experiments on benchmark datasets including EndoVis2022, AutoLapro, LapI2I and EndoVis2017 to verify the efficacy of SurgDepth. Specifically, SurgDepth achieves a new SOTA IoU of 0.86 on EndoVis 2022 SAR-RARP50 challenge and outperforms the current best method by at least 4%, using a shallow and compute efficient decoder consisting of ConvNeXt blocks.
Authors: Jiaxin Zhanga, Yunqin Lia, Tomohiro Fukudab, Bowen Wang
Abstract: Measuring urban safety perception is an important and complex task that traditionally relies heavily on human resources. This process often involves extensive field surveys, manual data collection, and subjective assessments, which can be time-consuming, costly, and sometimes inconsistent. Street View Images (SVIs), along with deep learning methods, provide a way to realize large-scale urban safety detection. However, achieving this goal often requires extensive human annotation to train safety ranking models, and the architectural differences between cities hinder the transferability of these models. Thus, a fully automated method for conducting safety evaluations is essential. Recent advances in multimodal large language models (MLLMs) have demonstrated powerful reasoning and analytical capabilities. Cutting-edge models, e.g., GPT-4 have shown surprising performance in many tasks. We employed these models for urban safety ranking on a human-annotated anchor set and validated that the results from MLLMs align closely with human perceptions. Additionally, we proposed a method based on the pre-trained Contrastive Language-Image Pre-training (CLIP) feature and K-Nearest Neighbors (K-NN) retrieval to quickly assess the safety index of the entire city. Experimental results show that our method outperforms existing training needed deep learning approaches, achieving efficient and accurate urban safety evaluations. The proposed automation for urban safety perception assessment is a valuable tool for city planners, policymakers, and researchers aiming to improve urban environments.
Authors: Sangjune Shin, Dongkun Shin
Abstract: Despite the rapid advancement of object detection algorithms, processing high-resolution images on embedded devices remains a significant challenge. Theoretically, the fully convolutional network architecture used in current real-time object detectors can handle all input resolutions. However, the substantial computational demands required to process high-resolution images render them impractical for real-time applications. To address this issue, real-time object detection models typically downsample the input image for inference, leading to a loss of detail and decreased accuracy. In response, we developed Octave-YOLO, designed to process high-resolution images in real-time within the constraints of embedded systems. We achieved this through the introduction of the cross frequency partial network (CFPNet), which divides the input feature map into low-resolution, low-frequency, and high-resolution, high-frequency sections. This configuration enables complex operations such as convolution bottlenecks and self-attention to be conducted exclusively on low-resolution feature maps while simultaneously preserving the details in high-resolution maps. Notably, this approach not only dramatically reduces the computational demands of convolution tasks but also allows for the integration of attention modules, which are typically challenging to implement in real-time applications, with minimal additional cost. Additionally, we have incorporated depthwise separable convolution into the core building blocks and downsampling layers to further decrease latency. Experimental results have shown that Octave-YOLO matches the performance of YOLOv8 while significantly reducing computational demands. For example, in 1080x1080 resolution, Octave-YOLO-N is 1.56 times faster than YOLOv8, achieving nearly the same accuracy on the COCO dataset with approximately 40 percent fewer parameters and FLOPs.
Authors: Tingzhang Luo, Mingxuan Du, Jiatao Shi, Xinxiang Chen, Bingchen Zhao, Shaoguang Huang
Abstract: This paper introduces a novel approach to Generalized Category Discovery (GCD) by leveraging the concept of contextuality to enhance the identification and classification of categories in unlabeled datasets. Drawing inspiration from human cognition's ability to recognize objects within their context, we propose a dual-context based method. Our model integrates two levels of contextuality: instance-level, where nearest-neighbor contexts are utilized for contrastive learning, and cluster-level, employing prototypical contrastive learning based on category prototypes. The integration of the contextual information effectively improves the feature learning and thereby the classification accuracy of all categories, which better deals with the real-world datasets. Different from the traditional semi-supervised and novel category discovery techniques, our model focuses on a more realistic and challenging scenario where both known and novel categories are present in the unlabeled data. Extensive experimental results on several benchmark data sets demonstrate that the proposed model outperforms the state-of-the-art. Code is available at: https://github.com/Clarence-CV/Contexuality-GCD
Authors: Chen Liu, Can Han, Chengfeng Zhou, Crystal Cai, Dahong Qian
Abstract: Gesture recognition based on surface electromyography (sEMG) has achieved significant progress in human-machine interaction (HMI). However, accurately recognizing predefined gestures within a closed set is still inadequate in practice; a robust open-set system needs to effectively reject unknown gestures while correctly classifying known ones. To handle this challenge, we first report prediction inconsistency discovered for unknown classes due to ensemble diversity, which can significantly facilitate the detection of unknown classes. Based on this insight, we propose an ensemble learning approach, PredIN, to explicitly magnify the prediction inconsistency by enhancing ensemble diversity. Specifically, PredIN maximizes the class feature distribution inconsistency among ensemble members to enhance diversity. Meanwhile, it optimizes inter-class separability within an individual ensemble member to maintain individual performance. Comprehensive experiments on various benchmark datasets demonstrate that the PredIN outperforms state-of-the-art methods by a clear margin.Our proposed method simultaneously achieves accurate closed-set classification for predefined gestures and effective rejection for unknown gestures, exhibiting its efficacy and superiority in open-set gesture recognition based on sEMG.
Authors: Wenjie Li, Heng Guo, Xuannan Liu, Kongming Liang, Jiani Hu, Zhanyu Ma, Jun Guo
Abstract: Face super-resolution aims to reconstruct a high-resolution face image from a low-resolution face image. Previous methods typically employ an encoder-decoder structure to extract facial structural features, where the direct downsampling inevitably introduces distortions, especially to high-frequency features such as edges. To address this issue, we propose a wavelet-based feature enhancement network, which mitigates feature distortion by losslessly decomposing the input feature into high and low-frequency components using the wavelet transform and processing them separately. To improve the efficiency of facial feature extraction, a full domain Transformer is further proposed to enhance local, regional, and global facial features. Such designs allow our method to perform better without stacking many modules as previous methods did. Experiments show that our method effectively balances performance, model size, and speed. Code link: https://github.com/PRIS-CV/WFEN.
Authors: Renke Wang, Meng Zhang, Jun Li, Jian Yan
Abstract: Generating high-fidelity garment animations through traditional workflows, from modeling to rendering, is both tedious and expensive. These workflows often require repetitive steps in response to updates in character motion, rendering viewpoint changes, or appearance edits. Although recent neural rendering offers an efficient solution for computationally intensive processes, it struggles with rendering complex garment animations containing fine wrinkle details and realistic garment-and-body occlusions, while maintaining structural consistency across frames and dense view rendering. In this paper, we propose a novel approach to directly synthesize garment animations from body motion sequences without the need for an explicit garment proxy. Our approach infers garment dynamic features from body motion, providing a preliminary overview of garment structure. Simultaneously, we capture detailed features from synthesized reference images of the garment's front and back, generated by a pre-trained image model. These features are then used to construct a neural radiance field that renders the garment animation video. Additionally, our technique enables garment recoloring by decomposing its visual elements. We demonstrate the generalizability of our method across unseen body motions and camera views, ensuring detailed structural consistency. Furthermore, we showcase its applicability to color editing on both real and synthetic garment data. Compared to existing neural rendering techniques, our method exhibits qualitative and quantitative improvements in garment dynamics and wrinkle detail modeling. Code is available at \url{https://github.com/wrk226/GarmentAnimationNeRF}.
Authors: Shohei Tanaka, Hao Wang, Yoshitaka Ushiku
Abstract: Scientific posters are used to present the contributions of scientific papers effectively in a graphical format. However, creating a well-designed poster that efficiently summarizes the core of a paper is both labor-intensive and time-consuming. A system that can automatically generate well-designed posters from scientific papers would reduce the workload of authors and help readers understand the outline of the paper visually. Despite the demand for poster generation systems, only a limited research has been conduced due to the lack of publicly available datasets. Thus, in this study, we built the SciPostLayout dataset, which consists of 7,855 scientific posters and manual layout annotations for layout analysis and generation. SciPostLayout also contains 100 scientific papers paired with the posters. All of the posters and papers in our dataset are under the CC-BY license and are publicly available. As benchmark tests for the collected dataset, we conducted experiments for layout analysis and generation utilizing existing computer vision models and found that both layout analysis and generation of posters using SciPostLayout are more challenging than with scientific papers. We also conducted experiments on generating layouts from scientific papers to demonstrate the potential of utilizing LLM as a scientific poster generation system. The dataset is publicly available at https://huggingface.co/datasets/omron-sinicx/scipostlayout_v2. The code is also publicly available at https://github.com/omron-sinicx/scipostlayout.
URLs: https://huggingface.co/datasets/omron-sinicx/scipostlayout_v2., https://github.com/omron-sinicx/scipostlayout.
Authors: Jinfan Hu, Jinjin Gu, Shiyao Yu, Fanghua Yu, Zheyuan Li, Zhiyuan You, Chaochao Lu, Chao Dong
Abstract: Deep neural networks have significantly improved the performance of low-level vision tasks but also increased the difficulty of interpretability. A deep understanding of deep models is beneficial for both network design and practical reliability. To take up this challenge, we introduce causality theory to interpret low-level vision models and propose a model-/task-agnostic method called Causal Effect Map (CEM). With CEM, we can visualize and quantify the input-output relationships on either positive or negative effects. After analyzing various low-level vision tasks with CEM, we have reached several interesting insights, such as: (1) Using more information of input images (e.g., larger receptive field) does NOT always yield positive outcomes. (2) Attempting to incorporate mechanisms with a global receptive field (e.g., channel attention) into image denoising may prove futile. (3) Integrating multiple tasks to train a general model could encourage the network to prioritize local information over global context. Based on the causal effect theory, the proposed diagnostic tool can refresh our common knowledge and bring a deeper understanding of low-level vision models. Codes are available at https://github.com/J-FHu/CEM.
Authors: Stefanos Gkikas, Manolis Tsiknakis
Abstract: Automatic pain assessment plays a critical role for advancing healthcare and optimizing pain management strategies. This study has been submitted to the First Multimodal Sensing Grand Challenge for Next-Gen Pain Assessment (AI4PAIN). The proposed multimodal framework utilizes facial videos and fNIRS and presents a modality-agnostic approach, alleviating the need for domain-specific models. Employing a dual ViT configuration and adopting waveform representations for the fNIRS, as well as for the extracted embeddings from the two modalities, demonstrate the efficacy of the proposed method, achieving an accuracy of 46.76% in the multilevel pain assessment task.
Authors: Stefanos Gkikas, Manolis Tsiknakis
Abstract: Pain assessment is essential in developing optimal pain management protocols to alleviate suffering and prevent functional decline in patients. Consequently, reliable and accurate automatic pain assessment systems are essential for continuous and effective patient monitoring. This study presents synthetic thermal videos generated by Generative Adversarial Networks integrated into the pain recognition pipeline and evaluates their efficacy. A framework consisting of a Vision-MLP and a Transformer-based module is utilized, employing RGB and synthetic thermal videos in unimodal and multimodal settings. Experiments conducted on facial videos from the BioVid database demonstrate the effectiveness of synthetic thermal videos and underline the potential advantages of it.
Authors: Artemis Llabr\'es, Arka Ujjal Dey, Dimosthenis Karatzas, Ernest Valveny
Abstract: We address the problem of detecting and mapping all books in a collection of images to entries in a given book catalogue. Instead of performing independent retrieval for each book detected, we treat the image-text mapping problem as a many-to-many matching process, looking for the best overall match between the two sets. We combine a state-of-the-art segmentation method (SAM) to detect book spines and extract book information using a commercial OCR. We then propose a two-stage approach for text-image matching, where CLIP embeddings are used first for fast matching, followed by a second slower stage to refine the matching, employing either the Hungarian Algorithm or a BERT-based model trained to cope with noisy OCR input and partial text matches. To evaluate our approach, we publish a new dataset of annotated bookshelf images that covers the whole book collection of a public library in Spain. In addition, we provide two target lists of book metadata, a closed-set of 15k book titles that corresponds to the known library inventory, and an open-set of 2.3M book titles to simulate an open-world scenario. We report results on two settings, on one hand on a matching-only task, where the book segments and OCR is given and the objective is to perform many-to-many matching against the target lists, and a combined detection and matching task, where books must be first detected and recognised before they are matched to the target list entries. We show that both the Hungarian Matching and the proposed BERT-based model outperform a fuzzy string matching baseline, and we highlight inherent limitations of the matching algorithms as the target increases in size, and when either of the two sets (detected books or target book list) is incomplete. The dataset and code are available at https://github.com/llabres/library-dataset
Authors: Guoliang Xu, Jianqin Yin, Feng Zhou, Yonghao Dang
Abstract: Previous methods usually only extract the image modality's information to recognize group activity. However, mining image information is approaching saturation, making it difficult to extract richer information. Therefore, extracting complementary information from other modalities to supplement image information has become increasingly important. In fact, action labels provide clear text information to express the action's semantics, which existing methods often overlook. Thus, we propose ActivityCLIP, a plug-and-play method for mining the text information contained in the action labels to supplement the image information for enhancing group activity recognition. ActivityCLIP consists of text and image branches, where the text branch is plugged into the image branch (The off-the-shelf image-based method). The text branch includes Image2Text and relation modeling modules. Specifically, we propose the knowledge transfer module, Image2Text, which adapts image information into text information extracted by CLIP via knowledge distillation. Further, to keep our method convenient, we add fewer trainable parameters based on the relation module of the image branch to model interaction relation in the text branch. To show our method's generality, we replicate three representative methods by ActivityCLIP, which adds only limited trainable parameters, achieving favorable performance improvements for each method. We also conduct extensive ablation studies and compare our method with state-of-the-art methods to demonstrate the effectiveness of ActivityCLIP.
Authors: Wenjun Huang, Jianguo Hu
Abstract: Multimodal Large Language Models (MLLMs) have attracted much attention due to their multifunctionality. However, traditional Transformer architectures incur significant overhead due to their secondary computational complexity. To address this issue, we introduce ML-Mamba, a multimodal language model that utilizes the latest and efficient Mamba-2 model for inference. Mamba-2 is known for its linear extension and fast processing of long sequences. We replace the Transformer based backbone with a pre-trained Mamba-2 model and explore methods for integrating 2D visual selective scanning mechanisms into multimodal learning. We also try various visual encoders and Mamba-2 model variants. Our extensive experiments conducted in various multimodal benchmark tests have demonstrated the competitive performance of ML-Mamba and highlighted the potential of state space models in multimodal tasks. The experimental results show that: (1) ML-Mamba achieves performance comparable to state-of-the-art methods such as TinyLaVA and MobileVLM v2 through its linear sequential modeling, while also having faster inference speed; (2) ML-Mamba performs well in visual hallucinations and spatial relationship judgment in closed set benchmark tests; (3) ML-Mamba achieves performance comparable to LLaVA while reducing the number of parameters by 40\%.(4) Compared to the multimodal model using the original Mamba model, the Mamba-2 based large-scale multimodal language model has stronger inference performance and effectiveness.
Authors: Diego Thomas, Briac Toussaint, Jean-Sebastien Franco, Edmond Boyer
Abstract: Volumetric shape representations have become ubiquitous in multi-view reconstruction tasks. They often build on regular voxel grids as discrete representations of 3D shape functions, such as SDF or radiance fields, either as the full shape model or as sampled instantiations of continuous representations, as with neural networks. Despite their proven efficiency, voxel representations come with the precision versus complexity trade-off. This inherent limitation can significantly impact performance when moving away from simple and uncluttered scenes. In this paper we investigate an alternative discretization strategy with the Centroidal Voronoi Tesselation (CVT). CVTs allow to better partition the observation space with respect to shape occupancy and to focus the discretization around shape surfaces. To leverage this discretization strategy for multi-view reconstruction, we introduce a volumetric optimization framework that combines explicit SDF fields with a shallow color network, in order to estimate 3D shape properties over tetrahedral grids. Experimental results with Chamfer statistics validate this approach with unprecedented reconstruction quality on various scenarios such as objects, open scenes or human.
Authors: Jihun Yi, Dahuin Jung, Sungroh Yoon
Abstract: The task of image anomaly detection (IAD) aims to identify deviations from normality in image data. These anomalies are patterns that deviate significantly from what the IAD model has learned from the data during training. However, in real-world scenarios, the criteria for what constitutes normality often change, necessitating the reclassification of previously anomalous instances as normal. To address this challenge, we propose a new scenario termed "normality addition," involving the post-training adjustment of decision boundaries to incorporate new normalities. To address this challenge, we propose a method called Normality Addition via Normality Detection (NAND), leveraging a vision-language model. NAND performs normality detection which detect patterns related to the intended normality within images based on textual descriptions. We then modify the results of a pre-trained IAD model to implement this normality addition. Using the benchmark dataset in IAD, MVTec AD, we establish an evaluation protocol for the normality addition task and empirically demonstrate the effectiveness of the NAND method.
Authors: Jiehui Tang, Xiaofei Wang, Zhen Xiao, Jiayi Liu, Xueliang Liu, Richang Hong
Abstract: This paper presents Team Xaiofei's innovative approach to exploring Face-Voice Association in Multilingual Environments (FAME) at ACM Multimedia 2024. We focus on the impact of different languages in face-voice matching by building upon Fusion and Orthogonal Projection (FOP), introducing four key components: a dual-branch structure, dynamic sample pair weighting, robust data augmentation, and score polarization strategy. Our dual-branch structure serves as an auxiliary mechanism to better integrate and provide more comprehensive information. We also introduce a dynamic weighting mechanism for various sample pairs to optimize learning. Data augmentation techniques are employed to enhance the model's generalization across diverse conditions. Additionally, score polarization strategy based on age and gender matching confidence clarifies and accentuates the final results. Our methods demonstrate significant effectiveness, achieving an equal error rate (EER) of 20.07 on the V2-EH dataset and 21.76 on the V1-EU dataset.
Authors: Sebastian N{\o}rgaard Llambias, Julia Machnio, Asbj{\o}rn Munk, Jakob Ambsdorf, Mads Nielsen, Mostafa Mehdipour Ghazi
Abstract: Medical image analysis using deep learning frameworks has advanced healthcare by automating complex tasks, but many existing frameworks lack flexibility, modularity, and user-friendliness. To address these challenges, we introduce Yucca, an open-source AI framework available at https://github.com/Sllambias/yucca, designed specifically for medical imaging applications and built on PyTorch and PyTorch Lightning. Yucca features a three-tiered architecture: Functional, Modules, and Pipeline, providing a comprehensive and customizable solution. Evaluated across diverse tasks such as cerebral microbleeds detection, white matter hyperintensity segmentation, and hippocampus segmentation, Yucca achieves state-of-the-art results, demonstrating its robustness and versatility. Yucca offers a powerful, flexible, and user-friendly platform for medical image analysis, inviting community contributions to advance its capabilities and impact.
Authors: Carlos Penarrubia, Jose J. Valero-Mas, Jorge Calvo-Zaragoza
Abstract: Text Recognition (TR) refers to the research area that focuses on retrieving textual information from images, a topic that has seen significant advancements in the last decade due to the use of Deep Neural Networks (DNN). However, these solutions often necessitate vast amounts of manually labeled or synthetic data. Addressing this challenge, Self-Supervised Learning (SSL) has gained attention by utilizing large datasets of unlabeled data to train DNN, thereby generating meaningful and robust representations. Although SSL was initially overlooked in TR because of its unique characteristics, recent years have witnessed a surge in the development of SSL methods specifically for this field. This rapid development, however, has led to many methods being explored independently, without taking previous efforts in methodology or comparison into account, thereby hindering progress in the field of research. This paper, therefore, seeks to consolidate the use of SSL in the field of TR, offering a critical and comprehensive overview of the current state of the art. We will review and analyze the existing methods, compare their results, and highlight inconsistencies in the current literature. This thorough analysis aims to provide general insights into the field, propose standardizations, identify new research directions, and foster its proper development.
Authors: Alexander Ponomarchuk, Ivan Kruzhilov, Galina Zubkova, Artem Shadrin, Ruslan Utegenov, Ivan Bessonov, Pavel Blinov
Abstract: The SYNTAX score has become a widely used measure of coronary disease severity , crucial in selecting the optimal mode of revascularization. This paper introduces a new medical regression and classification problem - automatically estimating SYNTAX score from coronary angiography. Our study presents a comprehensive dataset of 1,844 patients, featuring a balanced distribution of individuals with zero and non-zero scores. This dataset includes a first-of-its-kind, complete coronary angiography samples captured through a multi-view X-ray video, allowing one to observe coronary arteries from multiple perspectives. Furthermore, we present a novel, fully automatic end-to-end method for estimating the SYNTAX. For such a difficult task, we have achieved a solid coefficient of determination R2 of 0.51 in score predictions.
Authors: Takato Yasuno
Abstract: In regenerative medicine research, we experimentally design the composition of chemical medium. We add different components to 384-well plates and culture the biological cells. We monitor the condition of the cells and take time-lapse bioimages for morphological assay. In particular, precipitation can appear as artefacts in the image and contaminate the noise in the imaging assay. Inspecting precipitates is a tedious task for the observer, and differences in experience can lead to variations in judgement from person to person. The machine learning approach will remove the burden of human inspection and provide consistent inspection. In addition, precipitation features are as small as 10-20 {\mu}m. A 1200 pixel square well image resized under a resolution of 2.82 {\mu}m/pixel will result in a reduction in precipitation features. Dividing the well images into 240-pixel squares and learning without resizing preserves the resolution of the original image. In this study, we developed an application to automatically detect precipitation on 384-well plates utilising optical microscope images. We apply MN-pair contrastive clustering to extract precipitation classes from approximately 20,000 patch images. To detect precipitation features, we compare deeper FCDDs detectors with optional backbones and build a machine learning pipeline to detect precipitation from the maximum score of quadruplet well images using isolation Forest algorithm, where the anomaly score is ranged from zero to one. Furthermore, using this application we can visualise precipitation situ heatmap on a 384-well plate.
Authors: Yu Lu, Yuanzhi Liang, Linchao Zhu, Yi Yang
Abstract: Video diffusion models have made substantial progress in various video generation applications. However, training models for long video generation tasks require significant computational and data resources, posing a challenge to developing long video diffusion models. This paper investigates a straightforward and training-free approach to extend an existing short video diffusion model (e.g. pre-trained on 16-frame videos) for consistent long video generation (e.g. 128 frames). Our preliminary observation has found that directly applying the short video diffusion model to generate long videos can lead to severe video quality degradation. Further investigation reveals that this degradation is primarily due to the distortion of high-frequency components in long videos, characterized by a decrease in spatial high-frequency components and an increase in temporal high-frequency components. Motivated by this, we propose a novel solution named FreeLong to balance the frequency distribution of long video features during the denoising process. FreeLong blends the low-frequency components of global video features, which encapsulate the entire video sequence, with the high-frequency components of local video features that focus on shorter subsequences of frames. This approach maintains global consistency while incorporating diverse and high-quality spatiotemporal details from local videos, enhancing both the consistency and fidelity of long video generation. We evaluated FreeLong on multiple base video diffusion models and observed significant improvements. Additionally, our method supports coherent multi-prompt generation, ensuring both visual coherence and seamless transitions between scenes.
Authors: Benjamin Lambert, Florence Forbes, Senan Doyle, Michel Dojat
Abstract: Volumetry is one of the principal downstream applications of 3D medical image segmentation, for example, to detect abnormal tissue growth or for surgery planning. Conformal Prediction is a promising framework for uncertainty quantification, providing calibrated predictive intervals associated with automatic volume measurements. However, this methodology is based on the hypothesis that calibration and test samples are exchangeable, an assumption that is in practice often violated in medical image applications. A weighted formulation of Conformal Prediction can be framed to mitigate this issue, but its empirical investigation in the medical domain is still lacking. A potential reason is that it relies on the estimation of the density ratio between the calibration and test distributions, which is likely to be intractable in scenarios involving high-dimensional data. To circumvent this, we propose an efficient approach for density ratio estimation relying on the compressed latent representations generated by the segmentation model. Our experiments demonstrate the efficiency of our approach to reduce the coverage error in the presence of covariate shifts, in both synthetic and real-world settings. Our implementation is available at https://github.com/benolmbrt/wcp_miccai
Authors: Mingzhao Yang, Shangchao Su, Bin Li, Xiangyang Xue
Abstract: In recent years, the attention towards One-Shot Federated Learning (OSFL) has been driven by its capacity to minimize communication. With the development of the diffusion model (DM), several methods employ the DM for OSFL, utilizing model parameters, image features, or textual prompts as mediums to transfer the local client knowledge to the server. However, these mediums often require public datasets or the uniform feature extractor, significantly limiting their practicality. In this paper, we propose FedDEO, a Description-Enhanced One-Shot Federated Learning Method with DMs, offering a novel exploration of utilizing the DM in OSFL. The core idea of our method involves training local descriptions on the clients, serving as the medium to transfer the knowledge of the distributed clients to the server. Firstly, we train local descriptions on the client data to capture the characteristics of client distributions, which are then uploaded to the server. On the server, the descriptions are used as conditions to guide the DM in generating synthetic datasets that comply with the distributions of various clients, enabling the training of the aggregated model. Theoretical analyses and sufficient quantitation and visualization experiments on three large-scale real-world datasets demonstrate that through the training of local descriptions, the server is capable of generating synthetic datasets with high quality and diversity. Consequently, with advantages in communication and privacy protection, the aggregated model outperforms compared FL or diffusion-based OSFL methods and, on some clients, outperforms the performance ceiling of centralized training.
Authors: Chao Liu, Xin Liu, Zitong Yu, Yonghong Hou, Huanjing Yue, Jingyu Yang
Abstract: Deep neural networks (DNNs) have been applied in many computer vision tasks and achieved state-of-the-art (SOTA) performance. However, misclassification will occur when DNNs predict adversarial examples which are created by adding human-imperceptible adversarial noise to natural examples. This limits the application of DNN in security-critical fields. In order to enhance the robustness of models, previous research has primarily focused on the unimodal domain, such as image recognition and video understanding. Although multi-modal learning has achieved advanced performance in various tasks, such as action recognition, research on the robustness of RGB-skeleton action recognition models is scarce. In this paper, we systematically investigate how to improve the robustness of RGB-skeleton action recognition models. We initially conducted empirical analysis on the robustness of different modalities and observed that the skeleton modality is more robust than the RGB modality. Motivated by this observation, we propose the \formatword{A}ttention-based \formatword{M}odality \formatword{R}eweighter (\formatword{AMR}), which utilizes an attention layer to re-weight the two modalities, enabling the model to learn more robust features. Our AMR is plug-and-play, allowing easy integration with multimodal models. To demonstrate the effectiveness of AMR, we conducted extensive experiments on various datasets. For example, compared to the SOTA methods, AMR exhibits a 43.77\% improvement against PGD20 attacks on the NTU-RGB+D 60 dataset. Furthermore, it effectively balances the differences in robustness between different modalities.
Authors: Gagan Jain, Nidhi Hegde, Aditya Kusupati, Arsha Nagrani, Shyamal Buch, Prateek Jain, Anurag Arnab, Sujoy Paul
Abstract: The visual medium (images and videos) naturally contains a large amount of information redundancy, thereby providing a great opportunity for leveraging efficiency in processing. While Vision Transformer (ViT) based models scale effectively to large data regimes, they fail to capitalize on this inherent redundancy, leading to higher computational costs. Mixture of Experts (MoE) networks demonstrate scalability while maintaining same inference-time costs, but they come with a larger parameter footprint. We present Mixture of Nested Experts (MoNE), which utilizes a nested structure for experts, wherein individual experts fall on an increasing compute-accuracy curve. Given a compute budget, MoNE learns to dynamically choose tokens in a priority order, and thus redundant tokens are processed through cheaper nested experts. Using this framework, we achieve equivalent performance as the baseline models, while reducing inference time compute by over two-fold. We validate our approach on standard image and video datasets - ImageNet-21K, Kinetics400, and Something-Something-v2. We further highlight MoNE$'$s adaptability by showcasing its ability to maintain strong performance across different inference-time compute budgets on videos, using only a single trained model.
Authors: Hao Shu, Guo-Ping Qiu
Abstract: Image Edge detection (ED) is a base task in computer vision. While the performance of the ED algorithm has been improved greatly by introducing CNN-based models, current models still suffer from unsatisfactory precision rates especially when only a low error toleration distance is allowed. Therefore, model architecture for more precise predictions still needs an investigation. On the other hand, the unavoidable noise training data provided by humans would lead to unsatisfactory model predictions even when inputs are edge maps themselves, which also needs improvement. In this paper, more precise ED models are presented with cascaded skipping density blocks (CSDB). Our models obtain state-of-the-art(SOTA) predictions in several datasets, especially in average precision rate (AP), which is confirmed by extensive experiments. Moreover, our models do not include down-sample operations, demonstrating those widely believed operations are not necessary. Also, a novel modification on data augmentation for training is employed, which allows noiseless data to be employed in model training and thus improves the performance of models predicting on edge maps themselves.
Authors: Daniel Gallo Fern\'andez, R\u{a}zvan-Andrei Matisan, Alejandro Monroy Mu\~noz, Janusz Partyka
Abstract: Text-to-image generative models often present issues regarding fairness with respect to certain sensitive attributes, such as gender or skin tone. This study aims to reproduce the results presented in "ITI-GEN: Inclusive Text-to-Image Generation" by Zhang et al. (2023a), which introduces a model to improve inclusiveness in these kinds of models. We show that most of the claims made by the authors about ITI-GEN hold: it improves the diversity and quality of generated images, it is scalable to different domains, it has plug-and-play capabilities, and it is efficient from a computational point of view. However, ITI-GEN sometimes uses undesired attributes as proxy features and it is unable to disentangle some pairs of (correlated) attributes such as gender and baldness. In addition, when the number of considered attributes increases, the training time grows exponentially and ITI-GEN struggles to generate inclusive images for all elements in the joint distribution. To solve these issues, we propose using Hard Prompt Search with negative prompting, a method that does not require training and that handles negation better than vanilla Hard Prompt Search. Nonetheless, Hard Prompt Search (with or without negative prompting) cannot be used for continuous attributes that are hard to express in natural language, an area where ITI-GEN excels as it is guided by images during training. Finally, we propose combining ITI-GEN and Hard Prompt Search with negative prompting.
Authors: Dennis Vetter, Muhammad Ahsan, Diana Delicado, Thomas A. Neubauer, Thomas Wilke, Gemma Roig
Abstract: In this paper, we present our first proposal of a machine learning system for the classification of freshwater snails of the genus Radomaniola. We elaborate on the specific challenges encountered during system design, and how we tackled them; namely a small, very imbalanced dataset with a high number of classes and high visual similarity between classes. We then show how we employed triplet networks and the multiple input modalities of images, measurements, and genetic information to overcome these challenges and reach a performance comparable to that of a trained domain expert.
Authors: Delyan Boychev, Radostin Cholakov
Abstract: Generative models, such as diffusion models (DMs), variational autoencoders (VAEs), and generative adversarial networks (GANs), produce images with a level of authenticity that makes them nearly indistinguishable from real photos and artwork. While this capability is beneficial for many industries, the difficulty of identifying synthetic images leaves online media platforms vulnerable to impersonation and misinformation attempts. To support the development of defensive methods, we introduce ImagiNet, a high-resolution and balanced dataset for synthetic image detection, designed to mitigate potential biases in existing resources. It contains 200K examples, spanning four content categories: photos, paintings, faces, and uncategorized. Synthetic images are produced with open-source and proprietary generators, whereas real counterparts of the same content type are collected from public datasets. The structure of ImagiNet allows for a two-track evaluation system: i) classification as real or synthetic and ii) identification of the generative model. To establish a baseline, we train a ResNet-50 model using a self-supervised contrastive objective (SelfCon) for each track. The model demonstrates state-of-the-art performance and high inference speed across established benchmarks, achieving an AUC of up to 0.99 and balanced accuracy ranging from 86% to 95%, even under social network conditions that involve compression and resizing. Our data and code are available at https://github.com/delyan-boychev/imaginet.
Authors: Walid Bousselham, Sofian Chaybouti, Christian Rupprecht, Vittorio Ferrari, Hilde Kuehne
Abstract: Vision-language foundation models such as CLIP have achieved tremendous results in global vision-language alignment, but still show some limitations in creating representations for specific image regions. % To address this problem, we propose MaskInversion, a method that leverages the feature representations of pre-trained foundation models, such as CLIP, to generate a context-aware embedding for a query image region specified by a mask at test time. MaskInversion starts with initializing an embedding token and compares its explainability map, derived from the foundation model, to the query mask. The embedding token is then subsequently refined to approximate the query region by minimizing the discrepancy between its explainability map and the query mask. During this process, only the embedding vector is updated, while the underlying foundation model is kept frozen allowing to use MaskInversion with any pre-trained model. As deriving the explainability map involves computing its gradient, which can be expensive, we propose a gradient decomposition strategy that simplifies this computation. The learned region representation can be used for a broad range of tasks, including open-vocabulary class retrieval, referring expression comprehension, as well as for localized captioning and image generation. We evaluate the proposed method on all those tasks on several datasets such as PascalVOC, MSCOCO, RefCOCO, and OpenImagesV7 and show its capabilities compared to other SOTA approaches.
Authors: Chakkrit Termritthikun, Ayaz Umer, Suwichaya Suwanwimolkul, Feng Xia, Ivan Lee
Abstract: Recent advancements in deep convolutional neural networks have significantly improved the performance of saliency prediction. However, the manual configuration of the neural network architectures requires domain knowledge expertise and can still be time-consuming and error-prone. To solve this, we propose a new Neural Architecture Search (NAS) framework for saliency prediction with two contributions. Firstly, a supernet for saliency prediction is built with a weight-sharing network containing all candidate architectures, by integrating a dynamic convolution into the encoder-decoder in the supernet, termed SalNAS. Secondly, despite the fact that SalNAS is highly efficient (20.98 million parameters), it can suffer from the lack of generalization. To solve this, we propose a self-knowledge distillation approach, termed Self-KD, that trains the student SalNAS with the weighted average information between the ground truth and the prediction from the teacher model. The teacher model, while sharing the same architecture, contains the best-performing weights chosen by cross-validation. Self-KD can generalize well without the need to compute the gradient in the teacher model, enabling an efficient training system. By utilizing Self-KD, SalNAS outperforms other state-of-the-art saliency prediction models in most evaluation rubrics across seven benchmark datasets while being a lightweight model. The code will be available at https://github.com/chakkritte/SalNAS
Authors: Yimian Dai, Mengxuan Xiao, Yiming Zhu, Huan Wang, Kehua Guo, Jian Yang
Abstract: Infrared small target detection poses unique challenges due to the scarcity of intrinsic target features and the abundance of similar background distractors. We argue that background semantics play a pivotal role in distinguishing visually similar objects for this task. To address this, we introduce a new task -- clustered infrared small target detection, and present DenseSIRST, a novel benchmark dataset that provides per-pixel semantic annotations for background regions, enabling the transition from sparse to dense target detection. Leveraging this dataset, we propose the Background-Aware Feature Exchange Network (BAFE-Net), which transforms the detection paradigm from a single task focused on the foreground to a multi-task architecture that jointly performs target detection and background semantic segmentation. BAFE-Net introduces a cross-task feature hard-exchange mechanism to embed target and background semantics between the two tasks. Furthermore, we propose the Background-Aware Gaussian Copy-Paste (BAG-CP) method, which selectively pastes small targets into sky regions during training, avoiding the creation of false alarm targets in complex non-sky backgrounds. Extensive experiments validate the effectiveness of BAG-CP and BAFE-Net in improving target detection accuracy while reducing false alarms. The DenseSIRST dataset, code, and trained models are available at https://github.com/GrokCV/BAFE-Net.
Authors: Chaoqun Du, Yulin Wang, Jiayi Guo, Yizeng Han, Jie Zhou, Gao Huang
Abstract: Test-Time Adaptation (TTA) aims to adapt pre-trained models to the target domain during testing. In reality, this adaptability can be influenced by multiple factors. Researchers have identified various challenging scenarios and developed diverse methods to address these challenges, such as dealing with continual domain shifts, mixed domains, and temporally correlated or imbalanced class distributions. Despite these efforts, a unified and comprehensive benchmark has yet to be established. To this end, we propose a Unified Test-Time Adaptation (UniTTA) benchmark, which is comprehensive and widely applicable. Each scenario within the benchmark is fully described by a Markov state transition matrix for sampling from the original dataset. The UniTTA benchmark considers both domain and class as two independent dimensions of data and addresses various combinations of imbalance/balance and i.i.d./non-i.i.d./continual conditions, covering a total of \( (2 \times 3)^2 = 36 \) scenarios. It establishes a comprehensive evaluation benchmark for realistic TTA and provides a guideline for practitioners to select the most suitable TTA method. Alongside this benchmark, we propose a versatile UniTTA framework, which includes a Balanced Domain Normalization (BDN) layer and a COrrelated Feature Adaptation (COFA) method--designed to mitigate distribution gaps in domain and class, respectively. Extensive experiments demonstrate that our UniTTA framework excels within the UniTTA benchmark and achieves state-of-the-art performance on average. Our code is available at \url{https://github.com/LeapLabTHU/UniTTA}.
Authors: Jinmiao Zhao, Zelin Shi, Chuang Yu, Yunpeng Liu
Abstract: Recently, deep learning-based single-frame infrared small target (SIRST) detection technology has made significant progress. However, existing infrared small target detection methods are often optimized for a fixed image resolution, a single wavelength, or a specific imaging system, limiting their breadth and flexibility in practical applications. Therefore, we propose a refined infrared small target detection scheme based on an adjustable sensitivity (AS) strategy and multi-scale fusion. Specifically, a multi-scale model fusion framework based on multi-scale direction-aware network (MSDA-Net) is constructed, which uses input images of multiple scales to train multiple models and fuses them. Multi-scale fusion helps characterize the shape, edge, and texture features of the target from different scales, making the model more accurate and reliable in locating the target. At the same time, we fully consider the characteristics of the infrared small target detection task and construct an edge enhancement difficulty mining (EEDM) loss. The EEDM loss helps alleviate the problem of category imbalance and guides the network to pay more attention to difficult target areas and edge features during training. In addition, we propose an adjustable sensitivity strategy for post-processing. This strategy significantly improves the detection rate of infrared small targets while ensuring segmentation accuracy. Extensive experimental results show that the proposed scheme achieves the best performance. Notably, this scheme won the first prize in the PRCV 2024 wide-area infrared small target detection competition.
Authors: Keming Wu, Man Yao, Yuhong Chou, Xuerui Qiu, Rui Yang, Bo Xu, Guoqi Li
Abstract: Spiking Neural Networks (SNNs) have received widespread attention due to their unique neuronal dynamics and low-power nature. Previous research empirically shows that SNNs with Poisson coding are more robust than Artificial Neural Networks (ANNs) on small-scale datasets. However, it is still unclear in theory how the adversarial robustness of SNNs is derived, and whether SNNs can still maintain its adversarial robustness advantage on large-scale dataset tasks. This work theoretically demonstrates that SNN's inherent adversarial robustness stems from its Poisson coding. We reveal the conceptual equivalence of Poisson coding and randomized smoothing in defense strategies, and analyze in depth the trade-off between accuracy and adversarial robustness in SNNs via the proposed Randomized Smoothing Coding (RSC) method. Experiments demonstrate that the proposed RSC-SNNs show remarkable adversarial robustness, surpassing ANNs and achieving state-of-the-art robustness results on large-scale dataset ImageNet. Our open-source implementation code is available at this https URL: https://github.com/KemingWu/RSC-SNN.
Authors: Jing Yang, Runping Xi, Yingxin Lai, Xun Lin, Zitong Yu
Abstract: Diffusion-based personalized visual content generation technologies have achieved significant breakthroughs, allowing for the creation of specific objects by just learning from a few reference photos. However, when misused to fabricate fake news or unsettling content targeting individuals, these technologies could cause considerable societal harm. To address this problem, current methods generate adversarial samples by adversarially maximizing the training loss, thereby disrupting the output of any personalized generation model trained with these samples. However, the existing methods fail to achieve effective defense and maintain stealthiness, as they overlook the intrinsic properties of diffusion models. In this paper, we introduce a novel Dual-Domain Anti-Personalization framework (DDAP). Specifically, we have developed Spatial Perturbation Learning (SPL) by exploiting the fixed and perturbation-sensitive nature of the image encoder in personalized generation. Subsequently, we have designed a Frequency Perturbation Learning (FPL) method that utilizes the characteristics of diffusion models in the frequency domain. The SPL disrupts the overall texture of the generated images, while the FPL focuses on image details. By alternating between these two methods, we construct the DDAP framework, effectively harnessing the strengths of both domains. To further enhance the visual quality of the adversarial samples, we design a localization module to accurately capture attentive areas while ensuring the effectiveness of the attack and avoiding unnecessary disturbances in the background. Extensive experiments on facial benchmarks have shown that the proposed DDAP enhances the disruption of personalized generation models while also maintaining high quality in adversarial samples, making it more effective in protecting privacy in practical applications.
Authors: Wenxuan Wang, Quan Sun, Fan Zhang, Yepeng Tang, Jing Liu, Xinlong Wang
Abstract: Contrastive Language-Image Pre-training (CLIP), which excels at abstracting open-world representations across domains and modalities, has become a foundation for a variety of vision and multimodal tasks. However, recent studies reveal that CLIP has severe visual shortcomings, such as which can hardly distinguish orientation, quantity, color, structure, etc. These visual shortcomings also limit the perception capabilities of multimodal large language models (MLLMs) built on CLIP. The main reason could be that the image-text pairs used to train CLIP are inherently biased, due to the lack of the distinctiveness of the text and the diversity of images. In this work, we present a simple post-training approach for CLIP models, which largely overcomes its visual shortcomings via a self-supervised diffusion process. We introduce DIVA, which uses the DIffusion model as a Visual Assistant for CLIP. Specifically, DIVA leverages generative feedback from text-to-image diffusion models to optimize CLIP representations, with only images (without corresponding text). We demonstrate that DIVA improves CLIP's performance on the challenging MMVP-VLM benchmark which assesses fine-grained visual abilities to a large extent (e.g., 3-7%), and enhances the performance of MLLMs and vision models on multimodal understanding and segmentation tasks. Extensive evaluation on 29 image classification and retrieval benchmarks confirms that our framework preserves CLIP's strong zero-shot capabilities. The code will be available at https://github.com/baaivision/DIVA.
Authors: Xingchen Zeng, Haichuan Lin, Yilin Ye, Wei Zeng
Abstract: Emerging multimodal large language models (MLLMs) exhibit great potential for chart question answering (CQA). Recent efforts primarily focus on scaling up training datasets (i.e., charts, data tables, and question-answer (QA) pairs) through data collection and synthesis. However, our empirical study on existing MLLMs and CQA datasets reveals notable gaps. First, current data collection and synthesis focus on data volume and lack consideration of fine-grained visual encodings and QA tasks, resulting in unbalanced data distribution divergent from practical CQA scenarios. Second, existing work follows the training recipe of the base MLLMs initially designed for natural images, under-exploring the adaptation to unique chart characteristics, such as rich text elements. To fill the gap, we propose a visualization-referenced instruction tuning approach to guide the training dataset enhancement and model development. Specifically, we propose a novel data engine to effectively filter diverse and high-quality data from existing datasets and subsequently refine and augment the data using LLM-based generation techniques to better align with practical QA tasks and visual encodings. Then, to facilitate the adaptation to chart characteristics, we utilize the enriched data to train an MLLM by unfreezing the vision encoder and incorporating a mixture-of-resolution adaptation strategy for enhanced fine-grained recognition. Experimental results validate the effectiveness of our approach. Even with fewer training examples, our model consistently outperforms state-of-the-art CQA models on established benchmarks. We also contribute a dataset split as a benchmark for future research. Source codes and datasets of this paper are available at https://github.com/zengxingchen/ChartQA-MLLM.
Authors: Tuomas Varanka, Huai-Qian Khor, Yante Li, Mengting Wei, Hanwei Kung, Nicu Sebe, Guoying Zhao
Abstract: Generative models have surged in popularity recently due to their ability to produce high-quality images and video. However, steering these models to produce images with specific attributes and precise control remains challenging. Humans, particularly their faces, are central to content generation due to their ability to convey rich expressions and intent. Current generative models mostly generate flat neutral expressions and characterless smiles without authenticity. Other basic expressions like anger are possible, but are limited to the stereotypical expression, while other unconventional facial expressions like doubtful are difficult to reliably generate. In this work, we propose the use of AUs (action units) for facial expression control in face generation. AUs describe individual facial muscle movements based on facial anatomy, allowing precise and localized control over the intensity of facial movements. By combining different action units, we unlock the ability to create unconventional facial expressions that go beyond typical emotional models, enabling nuanced and authentic reactions reflective of real-world expressions. The proposed method can be seamlessly integrated with both text and image prompts using adapters, offering precise and intuitive control of the generated results. Code and dataset are available in {https://github.com/tvaranka/fineface}.
Authors: \c{C}a\u{g}han K\"oksal, Ghazal Ghazaei, Felix Holm, Azade Farshad, Nassir Navab
Abstract: Graph-based holistic scene representations facilitate surgical workflow understanding and have recently demonstrated significant success. However, this task is often hindered by the limited availability of densely annotated surgical scene data. In this work, we introduce an end-to-end framework for the generation and optimization of surgical scene graphs on a downstream task. Our approach leverages the flexibility of graph-based spectral clustering and the generalization capability of foundation models to generate unsupervised scene graphs with learnable properties. We reinforce the initial spatial graph with sparse temporal connections using local matches between consecutive frames to predict temporally consistent clusters across a temporal neighborhood. By jointly optimizing the spatiotemporal relations and node features of the dynamic scene graph with the downstream task of phase segmentation, we address the costly and annotation-burdensome task of semantic scene comprehension and scene graph generation in surgical videos using only weak surgical phase labels. Further, by incorporating effective intermediate scene representation disentanglement steps within the pipeline, our solution outperforms the SOTA on the CATARACTS dataset by 8% accuracy and 10% F1 score in surgical workflow recognition
Authors: Linfei Pan, D\'aniel Bar\'ath, Marc Pollefeys, Johannes L. Sch\"onberger
Abstract: Recovering 3D structure and camera motion from images has been a long-standing focus of computer vision research and is known as Structure-from-Motion (SfM). Solutions to this problem are categorized into incremental and global approaches. Until now, the most popular systems follow the incremental paradigm due to its superior accuracy and robustness, while global approaches are drastically more scalable and efficient. With this work, we revisit the problem of global SfM and propose GLOMAP as a new general-purpose system that outperforms the state of the art in global SfM. In terms of accuracy and robustness, we achieve results on-par or superior to COLMAP, the most widely used incremental SfM, while being orders of magnitude faster. We share our system as an open-source implementation at {https://github.com/colmap/glomap}.
Authors: Ray Zhang, Zheming Zhou, Min Sun, Omid Ghasemalizadeh, Cheng-Hao Kuo, Ryan Eustice, Maani Ghaffari, Arnie Sen
Abstract: This paper introduces a robust unsupervised SE(3) point cloud registration method that operates without requiring point correspondences. The method frames point clouds as functions in a reproducing kernel Hilbert space (RKHS), leveraging SE(3)-equivariant features for direct feature space registration. A novel RKHS distance metric is proposed, offering reliable performance amidst noise, outliers, and asymmetrical data. An unsupervised training approach is introduced to effectively handle limited ground truth data, facilitating adaptation to real datasets. The proposed method outperforms classical and supervised methods in terms of registration accuracy on both synthetic (ModelNet40) and real-world (ETH3D) noisy, outlier-rich datasets. To our best knowledge, this marks the first instance of successful real RGB-D odometry data registration using an equivariant method. The code is available at {https://sites.google.com/view/eccv24-equivalign}
Authors: Junyan Li, Delin Chen, Tianle Cai, Peihao Chen, Yining Hong, Zhenfang Chen, Yikang Shen, Chuang Gan
Abstract: Current high-resolution vision-language models encode images as high-resolution image tokens and exhaustively take all these tokens to compute attention, which significantly increases the computational cost. To address this problem, we propose FlexAttention, a flexible attention mechanism for efficient high-resolution vision-language models. Specifically, a high-resolution image is encoded both as high-resolution tokens and low-resolution tokens, where only the low-resolution tokens and a few selected high-resolution tokens are utilized to calculate the attention map, which greatly shrinks the computational cost. The high-resolution tokens are selected via a high-resolution selection module which could retrieve tokens of relevant regions based on an input attention map. The selected high-resolution tokens are then concatenated to the low-resolution tokens and text tokens, and input to a hierarchical self-attention layer which produces an attention map that could be used for the next-step high-resolution token selection. The hierarchical self-attention process and high-resolution token selection process are performed iteratively for each attention layer. Experiments on multimodal benchmarks prove that our FlexAttention outperforms existing high-resolution VLMs (e.g., relatively ~9% in V* Bench, ~7% in TextVQA), while also significantly reducing the computational cost by nearly 40%.
Authors: Yuanwen Yue, Anurag Das, Francis Engelmann, Siyu Tang, Jan Eric Lenssen
Abstract: Current visual foundation models are trained purely on unstructured 2D data, limiting their understanding of 3D structure of objects and scenes. In this work, we show that fine-tuning on 3D-aware data improves the quality of emerging semantic features. We design a method to lift semantic 2D features into an efficient 3D Gaussian representation, which allows us to re-render them for arbitrary views. Using the rendered 3D-aware features, we design a fine-tuning strategy to transfer such 3D awareness into a 2D foundation model. We demonstrate that models fine-tuned in that way produce features that readily improve downstream task performance in semantic segmentation and depth estimation through simple linear probing. Notably, though fined-tuned on a single indoor dataset, the improvement is transferable to a variety of indoor datasets and out-of-domain datasets. We hope our study encourages the community to consider injecting 3D awareness when training 2D foundation models. Project page: https://ywyue.github.io/FiT3D.
Authors: Ekaterina Iakovleva, Fabio Pizzati, Philip Torr, St\'ephane Lathuili\`ere
Abstract: Text-based editing diffusion models exhibit limited performance when the user's input instruction is ambiguous. To solve this problem, we propose $\textit{Specify ANd Edit}$ (SANE), a zero-shot inference pipeline for diffusion-based editing systems. We use a large language model (LLM) to decompose the input instruction into specific instructions, i.e. well-defined interventions to apply to the input image to satisfy the user's request. We benefit from the LLM-derived instructions along the original one, thanks to a novel denoising guidance strategy specifically designed for the task. Our experiments with three baselines and on two datasets demonstrate the benefits of SANE in all setups. Moreover, our pipeline improves the interpretability of editing models, and boosts the output diversity. We also demonstrate that our approach can be applied to any edit, whether ambiguous or not. Our code is public at https://github.com/fabvio/SANE.
Authors: Gautier Evennou, Vivien Chappelier, Ewa Kijak, Teddy Furon
Abstract: This paper proposes a novel approach towards image authentication and tampering detection by using watermarking as a communication channel for semantic information. We modify the HiDDeN deep-learning watermarking architecture to embed and extract high-dimensional real vectors representing image captions. Our method improves significantly robustness on both malign and benign edits. We also introduce a local confidence metric correlated with Message Recovery Rate, enhancing the method's practical applicability. This approach bridges the gap between traditional watermarking and passive forensic methods, offering a robust solution for image integrity verification.
Authors: Johnny Jingze Li, Vivek Kurien George, Gabriel A. Silva
Abstract: We introduce a novel yet straightforward neural network initialization scheme that modifies conventional methods like Xavier and Kaiming initialization. Inspired by the concept of emergence and leveraging the emergence measures proposed by Li (2023), our method adjusts the layer-wise weight scaling factors to achieve higher emergence values. This enhancement is easy to implement, requiring no additional optimization steps for initialization compared to GradInit. We evaluate our approach across various architectures, including MLP and convolutional architectures for image recognition, and transformers for machine translation. We demonstrate substantial improvements in both model accuracy and training speed, with and without batch normalization. The simplicity, theoretical innovation, and demonstrable empirical advantages of our method make it a potent enhancement to neural network initialization practices. These results suggest a promising direction for leveraging emergence to improve neural network training methodologies. Code is available at: https://github.com/johnnyjingzeli/EmergenceInit.
Authors: Tianyu Xiong, Skylar W. Wurster, Hanqi Guo, Tom Peterka, Han-Wei Shen
Abstract: Feature grid Scene Representation Networks (SRNs) have been applied to scientific data as compact functional surrogates for analysis and visualization. As SRNs are black-box lossy data representations, assessing the prediction quality is critical for scientific visualization applications to ensure that scientists can trust the information being visualized. Currently, existing architectures do not support inference time reconstruction quality assessment, as coordinate-level errors cannot be evaluated in the absence of ground truth data. We propose a parameter-efficient multi-decoder SRN (MDSRN) ensemble architecture consisting of a shared feature grid with multiple lightweight multi-layer perceptron decoders. MDSRN can generate a set of plausible predictions for a given input coordinate to compute the mean as the prediction of the multi-decoder ensemble and the variance as a confidence score. The coordinate-level variance can be rendered along with the data to inform the reconstruction quality, or be integrated into uncertainty-aware volume visualization algorithms. To prevent the misalignment between the quantified variance and the prediction quality, we propose a novel variance regularization loss for ensemble learning that promotes the Regularized multi-decoder SRN (RMDSRN) to obtain a more reliable variance that correlates closely to the true model error. We comprehensively evaluate the quality of variance quantification and data reconstruction of Monte Carlo Dropout, Mean Field Variational Inference, Deep Ensemble, and Predicting Variance compared to the proposed MDSRN and RMDSRN across diverse scalar field datasets. We demonstrate that RMDSRN attains the most accurate data reconstruction and competitive variance-error correlation among uncertain SRNs under the same neural network parameter budgets.
Authors: Srinidhi Hegde, Kaur Kullman, Thomas Grubb, Leslie Lait, Stephen Guimond, Matthias Zwicker
Abstract: Exploring scientific datasets with billions of samples in real-time visualization presents a challenge - balancing high-fidelity rendering with speed. This work introduces a novel renderer - Neural Accelerated Renderer (NAR), that uses the neural deferred rendering framework to visualize large-scale scientific point cloud data. NAR augments a real-time point cloud rendering pipeline with high-quality neural post-processing, making the approach ideal for interactive visualization at scale. Specifically, we train a neural network to learn the point cloud geometry from a high-performance multi-stream rasterizer and capture the desired postprocessing effects from a conventional high-quality renderer. We demonstrate the effectiveness of NAR by visualizing complex multidimensional Lagrangian flow fields and photometric scans of a large terrain and compare the renderings against the state-of-the-art high-quality renderers. Through extensive evaluation, we demonstrate that NAR prioritizes speed and scalability while retaining high visual fidelity. We achieve competitive frame rates of $>$ 126 fps for interactive rendering of $>$ 350M points (i.e., an effective throughput of $>$ 44 billion points per second) using $\sim$12 GB of memory on RTX 2080 Ti GPU. Furthermore, we show that NAR is generalizable across different point clouds with similar visualization needs and the desired post-processing effects could be obtained with substantial high quality even at lower resolutions of the original point cloud, further reducing the memory requirements.
Authors: Shikha Dubey, Yosep Chong, Beatrice Knudsen, Shireen Y. Elhabian
Abstract: This paper introduces a Virtual Immunohistochemistry Multiplex staining (VIMs) model designed to generate multiple immunohistochemistry (IHC) stains from a single hematoxylin and eosin (H&E) stained tissue section. IHC stains are crucial in pathology practice for resolving complex diagnostic questions and guiding patient treatment decisions. While commercial laboratories offer a wide array of up to 400 different antibody-based IHC stains, small biopsies often lack sufficient tissue for multiple stains while preserving material for subsequent molecular testing. This highlights the need for virtual IHC staining. Notably, VIMs is the first model to address this need, leveraging a large vision-language single-step diffusion model for virtual IHC multiplexing through text prompts for each IHC marker. VIMs is trained on uniplex paired H&E and IHC images, employing an adversarial training module. Testing of VIMs includes both paired and unpaired image sets. To enhance computational efficiency, VIMs utilizes a pre-trained large latent diffusion model fine-tuned with small, trainable weights through the Low-Rank Adapter (LoRA) approach. Experiments on nuclear and cytoplasmic IHC markers demonstrate that VIMs outperforms the base diffusion model and achieves performance comparable to Pix2Pix, a standard generative model for paired image translation. Multiple evaluation methods, including assessments by two pathologists, are used to determine the performance of VIMs. Additionally, experiments with different prompts highlight the impact of text conditioning. This paper represents the first attempt to accelerate histopathology research by demonstrating the generation of multiple IHC stains from a single H&E input using a single model trained solely on uniplex data.
Authors: Saige Rutherford, Thomas Wolfers, Charlotte Fraza, Nathaniel G. Harrnet, Christian F. Beckmann, Henricus G. Ruhe, Andre F. Marquand
Abstract: Reference classes in healthcare establish healthy norms, such as pediatric growth charts of height and weight, and are used to chart deviations from these norms which represent potential clinical risk. How the demographics of the reference class influence clinical interpretation of deviations is unknown. Using normative modeling, a method for building reference classes, we evaluate the fairness (racial bias) in reference models of structural brain images that are widely used in psychiatry and neurology. We test whether including race in the model creates fairer models. We predict self-reported race using the deviation scores from three different reference class normative models, to better understand bias in an integrated, multivariate sense. Across all of these tasks, we uncover racial disparities that are not easily addressed with existing data or commonly used modeling techniques. Our work suggests that deviations from the norm could be due to demographic mismatch with the reference class, and assigning clinical meaning to these deviations should be done with caution. Our approach also suggests that acquiring more representative samples is an urgent research priority.
Authors: Zunaira Rauf, Abdul Rehman Khan, Asifullah Khan
Abstract: Accurate nuclei segmentation is an essential foundation for various applications in computational pathology, including cancer diagnosis and treatment planning. Even slight variations in nuclei representations can significantly impact these downstream tasks. However, achieving accurate segmentation remains challenging due to factors like clustered nuclei, high intra-class variability in size and shape, resemblance to other cells, and color or contrast variations between nuclei and background. Despite the extensive utilization of Convolutional Neural Networks (CNNs) in medical image segmentation, they may have trouble capturing long-range dependencies crucial for accurate nuclei delineation. Transformers address this limitation but might miss essential low-level features. To overcome these limitations, we utilized CNN-Transformer-based techniques for nuclei segmentation in H&E stained histology images. In this work, we proposed two CNN-Transformer architectures, Nuclei Hybrid Vision Transformer (NucleiHVT) and Channel Boosted Nuclei Hybrid Vision Transformer (CB-NucleiHVT), that leverage the strengths of both CNNs and Transformers to effectively learn nuclei boundaries in multi-organ histology images. The first architecture, NucleiHVT is inspired by the UNet architecture and incorporates the dual attention mechanism to capture both multi-level and multi-scale context effectively. The CB-NucleiHVT network, on the other hand, utilizes the concept of channel boosting to learn diverse feature spaces, enhancing the model's ability to distinguish subtle variations in nuclei characteristics. Detailed evaluation of two medical image segmentation datasets shows that the proposed architectures outperform existing CNN-based, Transformer-based, and hybrid methods. The proposed networks demonstrated effective results both in terms of quantitative metrics, and qualitative visual assessment.
Authors: Bing Wang, Shengsheng Wang, Changchun Li, Renchu Guan, Ximing Li
Abstract: Nowadays, misinformation is widely spreading over various social media platforms and causes extremely negative impacts on society. To combat this issue, automatically identifying misinformation, especially those containing multimodal content, has attracted growing attention from the academic and industrial communities, and induced an active research topic named Multimodal Misinformation Detection (MMD). Typically, existing MMD methods capture the semantic correlation and inconsistency between multiple modalities, but neglect some potential clues in multimodal content. Recent studies suggest that manipulated traces of the images in articles are non-trivial clues for detecting misinformation. Meanwhile, we find that the underlying intentions behind the manipulation, e.g., harmful and harmless, also matter in MMD. Accordingly, in this work, we propose to detect misinformation by learning manipulation features that indicate whether the image has been manipulated, as well as intention features regarding the harmful and harmless intentions of the manipulation. Unfortunately, the manipulation and intention labels that make these features discriminative are unknown. To overcome the problem, we propose two weakly supervised signals as alternatives by introducing additional datasets on image manipulation detection and formulating two classification tasks as positive and unlabeled learning problems. Based on these ideas, we propose a novel MMD method, namely Harmfully Manipulated Images Matter in MMD (HAMI-M3D). Extensive experiments across three benchmark datasets can demonstrate that HAMI-M3D can consistently improve the performance of any MMD baselines.
Authors: Qihan Ren, Yang Xu, Junpeng Zhang, Yue Xin, Dongrui Liu, Quanshi Zhang
Abstract: This study proves the two-phase dynamics of a deep neural network (DNN) learning interactions. Despite the long disappointing view of the faithfulness of post-hoc explanation of a DNN, in recent years, a series of theorems have been proven to show that given an input sample, a small number of interactions between input variables can be considered as primitive inference patterns, which can faithfully represent every detailed inference logic of the DNN on this sample. Particularly, it has been observed that various DNNs all learn interactions of different complexities with two-phase dynamics, and this well explains how a DNN's generalization power changes from under-fitting to over-fitting. Therefore, in this study, we prove the dynamics of a DNN gradually encoding interactions of different complexities, which provides a theoretically grounded mechanism for the over-fitting of a DNN. Experiments show that our theory well predicts the real learning dynamics of various DNNs on different tasks.
Authors: Zhao Mingyang, Jia Xiaohong, Ma Lei, Shi Yuke, Jiang Jingen, Li Qizhai, Yan Dong-Ming, Huang Tiejun
Abstract: This work presents a novel and effective method for fitting multidimensional ellipsoids to scattered data in the contamination of noise and outliers. We approach the problem as a Bayesian parameter estimate process and maximize the posterior probability of a certain ellipsoidal solution given the data. We establish a more robust correlation between these points based on the predictive distribution within the Bayesian framework. We incorporate a uniform prior distribution to constrain the search for primitive parameters within an ellipsoidal domain, ensuring ellipsoid-specific results regardless of inputs. We then establish the connection between measurement point and model data via Bayes' rule to enhance the method's robustness against noise. Due to independent of spatial dimensions, the proposed method not only delivers high-quality fittings to challenging elongated ellipsoids but also generalizes well to multidimensional spaces. To address outlier disturbances, often overlooked by previous approaches, we further introduce a uniform distribution on top of the predictive distribution to significantly enhance the algorithm's robustness against outliers. We introduce an {\epsilon}-accelerated technique to expedite the convergence of EM considerably. To the best of our knowledge, this is the first comprehensive method capable of performing multidimensional ellipsoid specific fitting within the Bayesian optimization paradigm under diverse disturbances. We evaluate it across lower and higher dimensional spaces in the presence of heavy noise, outliers, and substantial variations in axis ratios. Also, we apply it to a wide range of practical applications such as microscopy cell counting, 3D reconstruction, geometric shape approximation, and magnetometer calibration tasks.
Authors: Peichao Li, Oscar MacCormac, Jonathan Shapey, Tom Vercauteren
Abstract: Hyperspectral imaging holds promises in surgical imaging by offering biological tissue differentiation capabilities with detailed information that is invisible to the naked eye. For intra-operative guidance, real-time spectral data capture and display is mandated. Snapshot mosaic hyperspectral cameras are currently seen as the most suitable technology given this requirement. However, snapshot mosaic imaging requires a demosaicking algorithm to fully restore the spatial and spectral details in the images. Modern demosaicking approaches typically rely on synthetic datasets to develop supervised learning methods, as it is practically impossible to simultaneously capture both snapshot and high-resolution spectral images of the exact same surgical scene. In this work, we present a self-supervised demosaicking and RGB reconstruction method that does not depend on paired high-resolution data as ground truth. We leverage unpaired standard high-resolution surgical microscopy images, which only provide RGB data but can be collected during routine surgeries. Adversarial learning complemented by self-supervised approaches are used to drive our hyperspectral-based RGB reconstruction into resembling surgical microscopy images and increasing the spatial resolution of our demosaicking. The spatial and spectral fidelity of the reconstructed hyperspectral images have been evaluated quantitatively. Moreover, a user study was conducted to evaluate the RGB visualisation generated from these spectral images. Both spatial detail and colour accuracy were assessed by neurosurgical experts. Our proposed self-supervised demosaicking method demonstrates improved results compared to existing methods, demonstrating its potential for seamless integration into intra-operative workflows.
Authors: Linkai Peng, Zheyuan Zhang, Gorkem Durak, Frank H. Miller, Alpay Medetalibeyoglu, Michael B. Wallace, Ulas Bagci
Abstract: Pancreatic cancer remains one of the leading causes of cancer-related mortality worldwide. Precise segmentation of pancreatic tumors from medical images is a bottleneck for effective clinical decision-making. However, achieving a high accuracy is often limited by the small size and availability of real patient data for training deep learning models. Recent approaches have employed synthetic data generation to augment training datasets. While promising, these methods may not yet meet the performance benchmarks required for real-world clinical use. This study critically evaluates the limitations of existing generative-AI based frameworks for pancreatic tumor segmentation. We conduct a series of experiments to investigate the impact of synthetic \textit{tumor size} and \textit{boundary definition} precision on model performance. Our findings demonstrate that: (1) strategically selecting a combination of synthetic tumor sizes is crucial for optimal segmentation outcomes, and (2) generating synthetic tumors with precise boundaries significantly improves model accuracy. These insights highlight the importance of utilizing refined synthetic data augmentation for enhancing the clinical utility of segmentation models in pancreatic cancer decision making including diagnosis, prognosis, and treatment plans. Our code will be available at https://github.com/lkpengcs/SynTumorAnalyzer.
Authors: Xin Zhao, Qianqian Zhu, Jialing Wu
Abstract: To address the challenges of similarity between lesions and surrounding tissues, overlapping appearances of partially benign and malignant nodules, and difficulty in classification, a deep learning network that integrates CNN and Transformer is proposed for the classification of benign and malignant breast lesions in ultrasound images. This network adopts a dual-branch architecture for local-global feature extraction, making full use of the advantages of CNN in extracting local features and the ability of ViT to extract global features to enhance the network's feature extraction capabilities for breast nodules. The local feature extraction branch employs a residual network with multiple attention-guided modules, which can effectively capture the local details and texture features of breast nodules, enhance sensitivity to subtle changes within the nodules, and thus can aid in accurate classification of their benign and malignancy. The global feature extraction branch utilizes the multi-head self-attention ViT network, which can capture the overall shape, boundary, and relationship with surrounding tissues, and thereby enhancing the understanding and modeling of both nodule and global image features. Experimental results on a public ultrasound breast nodule data set show that the proposed method is better than other comparison networks, This indicates that the fusion of CNN and Transformer networks can effectively improve the performance of the classification model and provide a powerful solution for the benign-malignant classification of ultrasound breast.
Authors: Al Mohimanul Islam, Sadia Shakiba Bhuiyan, Mysun Mashira, Md. Rayhan Ahmed, Salekul Islam, Swakkhar Shatabda
Abstract: Segmenting polyps in colonoscopy images is essential for the early identification and diagnosis of colorectal cancer, a significant cause of worldwide cancer deaths. Prior deep learning based models such as Attention based variation, UNet variations and Transformer-derived networks have had notable success in capturing intricate features and complex polyp shapes. In this study, we have introduced the DeepLabv3++ model which is an enhanced version of the DeepLabv3+ architecture. It is designed to improve the precision and robustness of polyp segmentation in colonoscopy images. We have utilized The proposed model incorporates diverse separable convolutional layers and attention mechanisms within the MSPP block, enhancing its capacity to capture multi-scale and directional features. Additionally, the redesigned decoder further transforms the extracted features from the encoder into a more meaningful segmentation map. Our model was evaluated on three public datasets (CVC-ColonDB, CVC-ClinicDB, Kvasir-SEG) achieving Dice coefficient scores of 96.20%, 96.54%, and 96.08%, respectively. The experimental analysis shows that DeepLabV3++ outperforms several state-of-the-art models in polyp segmentation tasks. Furthermore, compared to the baseline DeepLabV3+ model, our DeepLabV3++ with its MSPP module and redesigned decoder architecture, significantly reduced segmentation errors (e.g., false positives/negatives) across small, medium, and large polyps. This improvement in polyp delineation is crucial for accurate clinical decision-making in colonoscopy.
Authors: Shilin Xu, Xiangtai Li, Haobo Yuan, Lu Qi, Yunhai Tong, Ming-Hsuan Yang
Abstract: The recent surge in Multimodal Large Language Models (MLLMs) has showcased their remarkable potential for achieving generalized intelligence by integrating visual understanding into Large Language Models.Nevertheless, the sheer model size of MLLMs leads to substantial memory and computational demands that hinder their widespread deployment. In this work, we do not propose a new efficient model structure or train small-scale MLLMs from scratch. Instead, we focus on what matters for training small-scale MLLMs through knowledge distillation, which is the first step from the multimodal distillation perspective. Our extensive studies involve training strategies, model choices, and distillation algorithms in the knowledge distillation process. These results show that joint alignment for both tokens and logit alignment plays critical roles in teacher-student frameworks. In addition, we draw a series of intriguing observations from this study. By evaluating different benchmarks and proper strategy, even a 2.7B small-scale model can perform on par with larger models with 7B or 13B parameters. Our code and models will be publicly available for further research.
Authors: Mahiro Ukai, Shuhei Kurita, Atsushi Hashimoto, Yoshitaka Ushiku, Nakamasa Inoue
Abstract: Visual question answering aims to provide responses to natural language questions given visual input. Recently, visual programmatic models (VPMs), which generate executable programs to answer questions through large language models (LLMs), have attracted research interest. However, they often require long input prompts to provide the LLM with sufficient API usage details to generate relevant code. To address this limitation, we propose AdaCoder, an adaptive prompt compression framework for VPMs. AdaCoder operates in two phases: a compression phase and an inference phase. In the compression phase, given a preprompt that describes all API definitions in the Python language with example snippets of code, a set of compressed preprompts is generated, each depending on a specific question type. In the inference phase, given an input question, AdaCoder predicts the question type and chooses the appropriate corresponding compressed preprompt to generate code to answer the question. Notably, AdaCoder employs a single frozen LLM and pre-defined prompts, negating the necessity of additional training and maintaining adaptability across different powerful black-box LLMs such as GPT and Claude. In experiments, we apply AdaCoder to ViperGPT and demonstrate that it reduces token length by 71.1%, while maintaining or even improving the performance of visual question answering.
Authors: Weiliang Chen, Li Jia, Yang Zhou, Qianqian Ren
Abstract: Federated learning combined with blockchain empowers secure data sharing in autonomous driving applications. Nevertheless, with the increasing granularity and complexity of vehicle-generated data, the lack of data quality audits raises concerns about multi-party mistrust in trajectory prediction tasks. In response, this paper proposes an asynchronous federated learning data sharing method based on an interpretable reputation quantization mechanism utilizing graph neural network tools. Data providers share data structures under differential privacy constraints to ensure security while reducing redundant data. We implement deep reinforcement learning to categorize vehicles by reputation level, which optimizes the aggregation efficiency of federated learning. Experimental results demonstrate that the proposed data sharing scheme not only reinforces the security of the trajectory prediction task but also enhances prediction accuracy.
Authors: Stefanos Gkikas, Chariklia Chatzaki, Manolis Tsiknakis
Abstract: Pain is a complex phenomenon which is manifested and expressed by patients in various forms. The immediate and objective recognition of it is a great of importance in order to attain a reliable and unbiased healthcare system. In this work, we elaborate electrocardiography signals revealing the existence of variations in pain perception among different demographic groups. We exploit this insight by introducing a novel multi-task neural network for automatic pain estimation utilizing the age and the gender information of each individual, and show its advantages compared to other approaches.
Authors: Letian Shi, Qi Lv, Xiang Deng, Liqiang Nie
Abstract: In this technical report, we present our solution for the EgoPlan Challenge in ICML 2024. To address the real-world egocentric task planning problem, we introduce a novel planning framework which comprises three stages: long-term memory Extraction, context-awared Planning, and multi-iteration Decision, named EPD. Given the task goal, task progress, and current observation, the extraction model first extracts task-relevant memory information from the progress video, transforming the complex long video into summarized memory information. The planning model then combines the context of the memory information with fine-grained visual information from the current observation to predict the next action. Finally, through multi-iteration decision-making, the decision model comprehensively understands the task situation and current state to make the most realistic planning decision. On the EgoPlan-Test set, EPD achieves a planning accuracy of 53.85% over 1,584 egocentric task planning questions. We have made all codes available at https://github.com/Kkskkkskr/EPD .
Authors: Azmyin Md. Kamal, Nenyi K. N. Dadson, Donovan Gegg, Corina Barbalata
Abstract: In Monocular Keyframe Visual Simultaneous Localization and Mapping (MKVSLAM) frameworks, when incremental position tracking fails, global pose has to be recovered in a short-time window, also known as short-term relocalization. This capability is crucial for mobile robots to have reliable navigation, build accurate maps, and have precise behaviors around human collaborators. This paper focuses on the development of robust short-term relocalization capabilities for mobile robots using a monocular camera system. A novel multimodal keyframe descriptor is introduced, that contains semantic information of objects detected in the environment and the spatial information of the camera. Using this descriptor, a new Keyframe-based Place Recognition (KPR) method is proposed that is formulated as a multi-stage keyframe filtering algorithm, leading to a new relocalization pipeline for MKVSLAM systems. The proposed approach is evaluated over several indoor GPS denied datasets and demonstrates accurate pose recovery, in comparison to a bag-of-words approach.
Authors: Muhammad Arbab Arshad, Talukder Zaki Jubery, Tirtho Roy, Rim Nassiri, Asheesh K. Singh, Arti Singh, Chinmay Hegde, Baskar Ganapathysubramanian, Aditya Balu, Adarsh Krishnamurthy, Soumik Sarkar
Abstract: Plant stress phenotyping traditionally relies on expert assessments and specialized models, limiting scalability in agriculture. Recent advances in multimodal large language models (LLMs) offer potential solutions to this challenge. We present AgEval, a benchmark comprising 12 diverse plant stress phenotyping tasks, to evaluate these models' capabilities. Our study assesses zero-shot and few-shot in-context learning performance of state-of-the-art models, including Claude, GPT, Gemini, and LLaVA. Results show significant performance improvements with few-shot learning, with F1 scores increasing from 46.24% to 73.37% in 8-shot identification for the best-performing model. Few-shot examples from other classes in the dataset have negligible or negative impacts, although having the exact category example helps to increase performance by 15.38%. We also quantify the consistency of model performance across different classes within each task, finding that the coefficient of variance (CV) ranges from 26.02% to 58.03% across models, implying that subject matter expertise is needed - of 'difficult' classes - to achieve reliability in performance. AgEval establishes baseline metrics for multimodal LLMs in agricultural applications, offering insights into their promise for enhancing plant stress phenotyping at scale. Benchmark and code can be accessed at: https://anonymous.4open.science/r/AgEval/
Authors: Michael Livanos, Ian Davidson
Abstract: Deep anomaly detection (AD) is perhaps the most controversial of data analytic tasks as it identifies entities that are then specifically targeted for further investigation or exclusion. Also controversial is the application of AI to facial imaging data. This work explores the intersection of these two areas to understand two core questions: "Who" these algorithms are being unfair to and equally important "Why". Recent work has shown that deep AD can be unfair to different groups despite being unsupervised with a recent study showing that for portraits of people: men of color are far more likely to be chosen to be outliers. We study the two main categories of AD algorithms: autoencoder-based and single-class-based which effectively try to compress all the instances with those that can not be easily compressed being deemed to be outliers. We experimentally verify sources of unfairness such as the under-representation of a group (e.g. people of color are relatively rare), spurious group features (e.g. men are often photographed with hats), and group labeling noise (e.g. race is subjective). We conjecture that lack of compressibility is the main foundation and the others cause it but experimental results show otherwise and we present a natural hierarchy amongst them.
Authors: Yixuan Wu, Kaiyuan Hu, Qian Shao, Jintai Chen, Danny Z. Chen, Jian Wu
Abstract: The advent of telemedicine represents a transformative development in leveraging technology to extend the reach of specialized medical expertise to remote surgeries, a field where the immediacy of expert guidance is paramount. However, the intricate dynamics of Operating Room (OR) scene pose unique challenges for telemedicine, particularly in achieving high-fidelity, real-time scene reconstruction and transmission amidst obstructions and bandwidth limitations. This paper introduces TeleOR, a pioneering system designed to address these challenges through real-time OR scene reconstruction for Tele-intervention. TeleOR distinguishes itself with three innovative approaches: dynamic self-calibration, which leverages inherent scene features for calibration without the need for preset markers, allowing for obstacle avoidance and real-time camera adjustment; selective OR reconstruction, focusing on dynamically changing scene segments to reduce reconstruction complexity; and viewport-adaptive transmission, optimizing data transmission based on real-time client feedback to efficiently deliver high-quality 3D reconstructions within bandwidth constraints. Comprehensive experiments on the 4D-OR surgical scene dataset demostrate the superiority and applicability of TeleOR, illuminating the potential to revolutionize tele-interventions by overcoming the spatial and technical barriers inherent in remote surgical guidance.
Authors: Shivam Kumar, Samrat Chatterjee
Abstract: Predicting traits from images lacking visual cues is challenging, as algorithms are designed to capture visually correlated ground truth. This problem is critical in biomedical sciences, and their solution can improve the efficacy of non-invasive methods. For example, a recent challenge of predicting MGMT methylation status from MRI images is critical for treatment decisions of glioma patients. Using less robust models poses a significant risk in these critical scenarios and underscores the urgency of addressing this issue. Despite numerous efforts, contemporary models exhibit suboptimal performance, and underlying reasons for this limitation remain elusive. In this study, we demystify the complexity of MGMT status prediction through a comprehensive exploration by performing benchmarks of existing models adjoining transfer learning. Their architectures were further dissected by observing gradient flow across layers. Additionally, a feature selection strategy was applied to improve model interpretability. Our finding highlighted that current models are unlearnable and may require new architectures to explore applications in the real world. We believe our study will draw immediate attention and catalyse advancements in predictive modelling with non-visible cues.
Authors: Juhwan Choi, Junehyoung Kwon, JungMin Yun, Seunguk Yu, YoungBin Kim
Abstract: Domain generalizability is a crucial aspect of a deep learning model since it determines the capability of the model to perform well on data from unseen domains. However, research on the domain generalizability of deep learning models for vision-language tasks remains limited, primarily because of the lack of required datasets. To address these challenges, we propose VolDoGer: Vision-Language Dataset for Domain Generalization, a dedicated dataset designed for domain generalization that addresses three vision-language tasks: image captioning, visual question answering, and visual entailment. We constructed VolDoGer by extending LLM-based data annotation techniques to vision-language tasks, thereby alleviating the burden of recruiting human annotators. We evaluated the domain generalizability of various models, ranging from fine-tuned models to a recent multimodal large language model, through VolDoGer.
Authors: Tianhang Nan, Hao Quan, Yong Ding, Xingyu Li, Kai Yang, Xiaoyu Cui
Abstract: Multiple Instance Learning (MIL) has garnered widespread attention in the field of Whole Slide Image (WSI) classification as it replaces pixel-level manual annotation with diagnostic reports as labels, significantly reducing labor costs. Recent research has shown that bag-level MIL methods often yield better results because they can consider all patches of the WSI as a whole. However, a drawback of such methods is the incorporation of more redundant patches, leading to interference. To extract patches with high diagnostic value while excluding interfering patches to address this issue, we developed an attention-based feature distillation multi-instance learning (AFD-MIL) approach. This approach proposed the exclusion of redundant patches as a preprocessing operation in weakly supervised learning, directly mitigating interference from extensive noise. It also pioneers the use of attention mechanisms to distill features with high diagnostic value, as opposed to the traditional practice of indiscriminately and forcibly integrating all patches. Additionally, we introduced global loss optimization to finely control the feature distillation module. AFD-MIL is orthogonal to many existing MIL methods, leading to consistent performance improvements. This approach has surpassed the current state-of-the-art method, achieving 91.47% ACC (accuracy) and 94.29% AUC (area under the curve) on the Camelyon16 (Camelyon Challenge 2016, breast cancer), while 93.33% ACC and 98.17% AUC on the TCGA-NSCLC (The Cancer Genome Atlas Program: non-small cell lung cancer). Different feature distillation methods were used for the two datasets, tailored to the specific diseases, thereby improving performance and interpretability.
Authors: Tuan Van Vo, Minh Nhat Vu, Baoru Huang, An Vuong, Ngan Le, Thieu Vo, Anh Nguyen
Abstract: Grasp detection is an essential task in robotics with various industrial applications. However, traditional methods often struggle with occlusions and do not utilize language for grasping. Incorporating natural language into grasp detection remains a challenging task and largely unexplored. To address this gap, we propose a new method for language-driven grasp detection with mask-guided attention by utilizing the transformer attention mechanism with semantic segmentation features. Our approach integrates visual data, segmentation mask features, and natural language instructions, significantly improving grasp detection accuracy. Our work introduces a new framework for language-driven grasp detection, paving the way for language-driven robotic applications. Intensive experiments show that our method outperforms other recent baselines by a clear margin, with a 10.0% success score improvement. We further validate our method in real-world robotic experiments, confirming the effectiveness of our approach.
Authors: Jacob Sam, Karan Patel, Mike Saad
Abstract: In the field of architecture, the conversion of single images into 2 and 1/2D and 3D meshes is a promising technology that enhances design visualization and efficiency. This paper evaluates four innovative methods: "One-2-3-45," "CRM: Single Image to 3D Textured Mesh with Convolutional Reconstruction Model," "Instant Mesh," and "Image-to-Mesh." These methods are at the forefront of this technology, focusing on their applicability in architectural design and visualization. They streamline the creation of 3D architectural models, enabling rapid prototyping and detailed visualization from minimal initial inputs, such as photographs or simple sketches.One-2-3-45 leverages a diffusion-based approach to generate multi-view reconstructions, ensuring high geometric fidelity and texture quality. CRM utilizes a convolutional network to integrate geometric priors into its architecture, producing detailed and textured meshes quickly and efficiently. Instant Mesh combines the strengths of multi-view diffusion and sparse-view models to offer speed and scalability, suitable for diverse architectural projects. Image-to-Mesh leverages a generative adversarial network (GAN) to produce 3D meshes from single images, focusing on maintaining high texture fidelity and geometric accuracy by incorporating image and depth map data into its training process. It uses a hybrid approach that combines voxel-based representations with surface reconstruction techniques to ensure detailed and realistic 3D models.This comparative study highlights each method's contribution to reducing design cycle times, improving accuracy, and enabling flexible adaptations to various architectural styles and requirements. By providing architects with powerful tools for rapid visualization and iteration, these advancements in 3D mesh generation are set to revolutionize architectural practices.
Authors: Kanghyun Choi, Hye Yoon Lee, Dain Kwon, SunJong Park, Kyuyeun Kim, Noseong Park, Jinho Lee
Abstract: Data-free quantization (DFQ) is a technique that creates a lightweight network from its full-precision counterpart without the original training data, often through a synthetic dataset. Although several DFQ methods have been proposed for vision transformer (ViT) architectures, they fail to achieve efficacy in low-bit settings. Examining the existing methods, we identify that their synthetic data produce misaligned attention maps, while those of the real samples are highly aligned. From the observation of aligned attention, we find that aligning attention maps of synthetic data helps to improve the overall performance of quantized ViTs. Motivated by this finding, we devise \aname, a novel DFQ method designed for ViTs that focuses on inter-head attention similarity. First, we generate synthetic data by aligning head-wise attention responses in relation to spatial query patches. Then, we apply head-wise structural attention distillation to align the attention maps of the quantized network to those of the full-precision teacher. The experimental results show that the proposed method significantly outperforms baselines, setting a new state-of-the-art performance for data-free ViT quantization.
Authors: Fangyijie Wang, Gu\'enol\'e Silvestre, Kathleen M. Curran
Abstract: Accurate measurement of fetal head circumference is crucial for estimating fetal growth during routine prenatal screening. Prior to measurement, it is necessary to accurately identify and segment the region of interest, specifically the fetal head, in ultrasound images. Recent advancements in deep learning techniques have shown significant progress in segmenting the fetal head using encoder-decoder models. Among these models, U-Net has become a standard approach for accurate segmentation. However, training an encoder-decoder model can be a time-consuming process that demands substantial computational resources. Moreover, fine-tuning these models is particularly challenging when there is a limited amount of data available. There are still no "best-practice" guidelines for optimal fine-tuning of U-net for fetal ultrasound image segmentation. This work summarizes existing fine-tuning strategies with various backbone architectures, model components, and fine-tuning strategies across ultrasound data from Netherlands, Spain, Malawi, Egypt and Algeria. Our study shows that (1) fine-tuning U-Net leads to better performance than training from scratch, (2) fine-tuning strategies in decoder are superior to other strategies, (3) network architecture with less number of parameters can achieve similar or better performance. We also demonstrate the effectiveness of fine-tuning strategies in low-resource settings and further expand our experiments into few-shot learning. Lastly, we publicly released our code and specific fine-tuned weights.
Authors: Ruochen Li, Jiazhen Pan, Youxiang Zhu, Juncheng Ni, Daniel Rueckert
Abstract: Cardiac Magnetic Resonance Imaging (CMR) is the gold standard for diagnosing cardiovascular diseases. Clinical diagnoses predominantly rely on magnitude-only Digital Imaging and Communications in Medicine (DICOM) images, omitting crucial phase information that might provide additional diagnostic benefits. In contrast, k-space is complex-valued and encompasses both magnitude and phase information, while humans cannot directly perceive. In this work, we propose KMAE, a Transformer-based model specifically designed to process k-space data directly, eliminating conventional intermediary conversion steps to the image domain. KMAE can handle critical cardiac disease classification, relevant phenotype regression, and cardiac morphology segmentation tasks. We utilize this model to investigate the potential of k-space-based diagnosis in cardiac MRI. Notably, this model achieves competitive classification and regression performance compared to image-domain methods e.g. Masked Autoencoders (MAEs) and delivers satisfactory segmentation performance with a myocardium dice score of 0.884. Last but not least, our model exhibits robust performance with consistent results even when the k-space is 8* undersampled. We encourage the MR community to explore the untapped potential of k-space and pursue end-to-end, automated diagnosis with reduced human intervention.
Authors: Mikel Williams-Lekuona, Georgina Cosma
Abstract: In the field of Image-Text Retrieval (ITR), recent advancements have leveraged large-scale Vision-Language Pretraining (VLP) for Fine-Grained (FG) instance-level retrieval, achieving high accuracy at the cost of increased computational complexity. For Coarse-Grained (CG) category-level retrieval, prominent approaches employ Cross-Modal Hashing (CMH) to prioritise efficiency, albeit at the cost of retrieval performance. Due to differences in methodologies, FG and CG models are rarely compared directly within evaluations in the literature, resulting in a lack of empirical data quantifying the retrieval performance-efficiency tradeoffs between the two. This paper addresses this gap by introducing the \texttt{FiCo-ITR} library, which standardises evaluation methodologies for both FG and CG models, facilitating direct comparisons. We conduct empirical evaluations of representative models from both subfields, analysing precision, recall, and computational complexity across varying data scales. Our findings offer new insights into the performance-efficiency trade-offs between recent representative FG and CG models, highlighting their respective strengths and limitations. These findings provide the foundation necessary to make more informed decisions regarding model selection for specific retrieval tasks and highlight avenues for future research into hybrid systems that leverage the strengths of both FG and CG approaches.
Authors: Zhenqi He, Wenrui Liu, Minghao Yin, Kai Han
Abstract: Histological artifacts pose challenges for both pathologists and Computer-Aided Diagnosis (CAD) systems, leading to errors in analysis. Current approaches for histological artifact restoration, based on Generative Adversarial Networks (GANs) and pixel-level Diffusion Models, suffer from performance limitations and computational inefficiencies. In this paper, we propose a novel framework, LatentArtiFusion, which leverages the latent diffusion model (LDM) to reconstruct histological artifacts with high performance and computational efficiency. Unlike traditional pixel-level diffusion frameworks, LatentArtiFusion executes the restoration process in a lower-dimensional latent space, significantly improving computational efficiency. Moreover, we introduce a novel regional artifact reconstruction algorithm in latent space to prevent mistransfer in non-artifact regions, distinguishing our approach from GAN-based methods. Through extensive experiments on real-world histology datasets, LatentArtiFusion demonstrates remarkable speed, outperforming state-of-the-art pixel-level diffusion frameworks by more than 30X. It also consistently surpasses GAN-based methods by at least 5% across multiple evaluation metrics. Furthermore, we evaluate the effectiveness of our proposed framework in downstream tissue classification tasks, showcasing its practical utility. Code is available at https://github.com/bugs-creator/LatentArtiFusion.
Authors: Jinghuan Shang, Karl Schmeckpeper, Brandon B. May, Maria Vittoria Minniti, Tarik Kelestemur, David Watkins, Laura Herlant
Abstract: Vision-based robot policy learning, which maps visual inputs to actions, necessitates a holistic understanding of diverse visual tasks beyond single-task needs like classification or segmentation. Inspired by this, we introduce Theia, a vision foundation model for robot learning that distills multiple off-the-shelf vision foundation models trained on varied vision tasks. Theia's rich visual representations encode diverse visual knowledge, enhancing downstream robot learning. Extensive experiments demonstrate that Theia outperforms its teacher models and prior robot learning models using less training data and smaller model sizes. Additionally, we quantify the quality of pre-trained visual representations and hypothesize that higher entropy in feature norm distributions leads to improved robot learning performance. Code and models are available at https://github.com/bdaiinstitute/theia.
Authors: Jian Wang, Razieh Faghihpirayesh, Polina Golland, Ali Ghoulipour
Abstract: In this paper, we introduce SpaER, a pioneering method for fetal motion tracking that leverages equivariant filters and self-attention mechanisms to effectively learn spatio-temporal representations. Different from conventional approaches that statically estimate fetal brain motions from pairs of images, our method dynamically tracks the rigid movement patterns of the fetal head across temporal and spatial dimensions. Specifically, we first develop an equivariant neural network that efficiently learns rigid motion sequences through low-dimensional spatial representations of images. Subsequently, we learn spatio-temporal representations by incorporating time encoding and self-attention neural network layers. This approach allows for the capture of long-term dependencies of fetal brain motion and addresses alignment errors due to contrast changes and severe motion artifacts. Our model also provides a geometric deformation estimation that properly addresses image distortions among all time frames. To the best of our knowledge, our approach is the first to learn spatial-temporal representations via deep neural networks for fetal motion tracking without data augmentation. We validated our model using real fetal echo-planar images with simulated and real motions. Our method carries significant potential value in accurately measuring, tracking, and correcting fetal motion in fetal MRI sequences.
Authors: Yiming Huang, Beilei Cui, Ikemura Kei, Jiekai Zhang, Long Bai, Hongliang Ren
Abstract: The recent advance in neural rendering has enabled the ability to reconstruct high-quality 4D scenes using neural networks. Although 4D neural reconstruction is popular, registration for such representations remains a challenging task, especially for dynamic scene registration in surgical planning and simulation. In this paper, we propose a novel strategy for dynamic surgical neural scene registration. We first utilize 4D Gaussian Splatting to represent the surgical scene and capture both static and dynamic scenes effectively. Then, a spatial aware feature aggregation method, Spatially Weight Cluttering (SWC) is proposed to accurately align the feature between surgical scenes, enabling precise and realistic surgical simulations. Lastly, we present a novel strategy of deformable scene registration to register two dynamic scenes. By incorporating both spatial and temporal information for correspondence matching, our approach achieves superior performance compared to existing registration methods for implicit neural representation. The proposed method has the potential to improve surgical planning and training, ultimately leading to better patient outcomes.
Authors: Jayesh Singla, Ananye Agarwal, Deepak Pathak
Abstract: Despite extreme sample inefficiency, on-policy reinforcement learning, aka policy gradients, has become a fundamental tool in decision-making problems. With the recent advances in GPU-driven simulation, the ability to collect large amounts of data for RL training has scaled exponentially. However, we show that current RL methods, e.g. PPO, fail to ingest the benefit of parallelized environments beyond a certain point and their performance saturates. To address this, we propose a new on-policy RL algorithm that can effectively leverage large-scale environments by splitting them into chunks and fusing them back together via importance sampling. Our algorithm, termed SAPG, shows significantly higher performance across a variety of challenging environments where vanilla PPO and other strong baselines fail to achieve high performance. Website at https://sapg-rl.github.io/
Authors: Rodney Lalonde, Naji Khosravan, Ulas Bagci
Abstract: Capsule networks promise significant benefits over convolutional networks by storing stronger internal representations, and routing information based on the agreement between intermediate representations' projections. Despite this, their success has been limited to small-scale classification datasets due to their computationally expensive nature. Though memory efficient, convolutional capsules impose geometric constraints that fundamentally limit the ability of capsules to model the pose/deformation of objects. Further, they do not address the bigger memory concern of class-capsules scaling up to bigger tasks such as detection or large-scale classification. In this study, we introduce a new family of capsule networks, deformable capsules (\textit{DeformCaps}), to address a very important problem in computer vision: object detection. We propose two new algorithms associated with our \textit{DeformCaps}: a novel capsule structure (\textit{SplitCaps}), and a novel dynamic routing algorithm (\textit{SE-Routing}), which balance computational efficiency with the need for modeling a large number of objects and classes, which have never been achieved with capsule networks before. We demonstrate that the proposed methods efficiently scale up to create the first-ever capsule network for object detection in the literature. Our proposed architecture is a one-stage detection framework and it obtains results on MS COCO which are on par with state-of-the-art one-stage CNN-based methods, while producing fewer false positive detection, generalizing to unusual poses/viewpoints of objects.
Authors: Xiaoyan Xing, Yanlin Qian, Sibo Feng, Yuhan Dong, Jiri Matas
Abstract: In this paper, we present Point Cloud Color Constancy, in short PCCC, an illumination chromaticity estimation algorithm exploiting a point cloud. We leverage the depth information captured by the time-of-flight (ToF) sensor mounted rigidly with the RGB sensor, and form a 6D cloud where each point contains the coordinates and RGB intensities, noted as (x,y,z,r,g,b). PCCC applies the PointNet architecture to the color constancy problem, deriving the illumination vector point-wise and then making a global decision about the global illumination chromaticity. On two popular RGB-D datasets, which we extend with illumination information, as well as on a novel benchmark, PCCC obtains lower error than the state-of-the-art algorithms. Our method is simple and fast, requiring merely 16*16-size input and reaching speed over 500 fps, including the cost of building the point cloud and net inference.
Authors: Dongzhuo Li
Abstract: Deep generative models such as GANs, normalizing flows, and diffusion models are powerful regularizers for inverse problems. They exhibit great potential for helping reduce ill-posedness and attain high-quality results. However, the latent tensors of such deep generative models can fall out of the desired high-dimensional standard Gaussian distribution during inversion, particularly in the presence of data noise and inaccurate forward models, leading to low-fidelity solutions. To address this issue, we propose to reparameterize and Gaussianize the latent tensors using novel differentiable data-dependent layers wherein custom operators are defined by solving optimization problems. These proposed layers constrain inverse problems to obtain high-fidelity in-distribution solutions. We validate our technique on three inversion tasks: compressive-sensing MRI, image deblurring, and eikonal tomography (a nonlinear PDE-constrained inverse problem) using two representative deep generative models: StyleGAN2 and Glow. Our approach achieves state-of-the-art performance in terms of accuracy and consistency.
Authors: Yunhao Ba, Howard Zhang, Ethan Yang, Akira Suzuki, Arnold Pfahnl, Chethan Chinder Chandrappa, Celso de Melo, Suya You, Stefano Soatto, Alex Wong, Achuta Kadambi
Abstract: We propose a large-scale dataset of real-world rainy and clean image pairs and a method to remove degradations, induced by rain streaks and rain accumulation, from the image. As there exists no real-world dataset for deraining, current state-of-the-art methods rely on synthetic data and thus are limited by the sim2real domain gap; moreover, rigorous evaluation remains a challenge due to the absence of a real paired dataset. We fill this gap by collecting a real paired deraining dataset through meticulous control of non-rain variations. Our dataset enables paired training and quantitative evaluation for diverse real-world rain phenomena (e.g. rain streaks and rain accumulation). To learn a representation robust to rain phenomena, we propose a deep neural network that reconstructs the underlying scene by minimizing a rain-robust loss between rainy and clean images. Extensive experiments demonstrate that our model outperforms the state-of-the-art deraining methods on real rainy images under various conditions. Project website: https://visual.ee.ucla.edu/gt_rain.htm/.
Authors: Shuai Zhao, Linchao Zhu, Xiaohan Wang, Yi Yang
Abstract: Self-supervised learning makes significant progress in pre-training large models, but struggles with small models. Mainstream solutions to this problem rely mainly on knowledge distillation, which involves a two-stage procedure: first training a large teacher model and then distilling it to improve the generalization ability of smaller ones. In this work, we introduce another one-stage solution to obtain pre-trained small models without the need for extra teachers, namely, slimmable networks for contrastive self-supervised learning (SlimCLR). A slimmable network consists of a full network and several weight-sharing sub-networks, which can be pre-trained once to obtain various networks, including small ones with low computation costs. However, interference between weight-sharing networks leads to severe performance degradation in self-supervised cases, as evidenced by gradient magnitude imbalance and gradient direction divergence. The former indicates that a small proportion of parameters produce dominant gradients during backpropagation, while the main parameters may not be fully optimized. The latter shows that the gradient direction is disordered, and the optimization process is unstable. To address these issues, we introduce three techniques to make the main parameters produce dominant gradients and sub-networks have consistent outputs. These techniques include slow start training of sub-networks, online distillation, and loss re-weighting according to model sizes. Furthermore, theoretical results are presented to demonstrate that a single slimmable linear layer is sub-optimal during linear evaluation. Thus a switchable linear probe layer is applied during linear evaluation. We instantiate SlimCLR with typical contrastive learning frameworks and achieve better performance than previous arts with fewer parameters and FLOPs. The code is at https://github.com/mzhaoshuai/SlimCLR.
Authors: Yuecong Xu, Haozhi Cao, Zhenghua Chen, Xiaoli Li, Lihua Xie, Jianfei Yang
Abstract: Video analysis tasks such as action recognition have received increasing research interest with growing applications in fields such as smart healthcare, thanks to the introduction of large-scale datasets and deep learning-based representations. However, video models trained on existing datasets suffer from significant performance degradation when deployed directly to real-world applications due to domain shifts between the training public video datasets (source video domains) and real-world videos (target video domains). Further, with the high cost of video annotation, it is more practical to use unlabeled videos for training. To tackle performance degradation and address concerns in high video annotation cost uniformly, the video unsupervised domain adaptation (VUDA) is introduced to adapt video models from the labeled source domain to the unlabeled target domain by alleviating video domain shift, improving the generalizability and portability of video models. This paper surveys recent progress in VUDA with deep learning. We begin with the motivation of VUDA, followed by its definition, and recent progress of methods for both closed-set VUDA and VUDA under different scenarios, and current benchmark datasets for VUDA research. Eventually, future directions are provided to promote further VUDA research. The repository of this survey is provided at https://github.com/xuyu0010/awesome-video-domain-adaptation.
URLs: https://github.com/xuyu0010/awesome-video-domain-adaptation.
Authors: Mengqi Guo, Chen Li, Hanlin Chen, Gim Hee Lee
Abstract: Recent neural implicit representations (NIRs) have achieved great success in the tasks of 3D reconstruction and novel view synthesis. However, they require the images of a scene from different camera views to be available for one-time training. This is expensive especially for scenarios with large-scale scenes and limited data storage. In view of this, we explore the task of incremental learning for NIRs in this work. We design a student-teacher framework to mitigate the catastrophic forgetting problem. Specifically, we iterate the process of using the student as the teacher at the end of each time step and let the teacher guide the training of the student in the next step. As a result, the student network is able to learn new information from the streaming data and retain old knowledge from the teacher network simultaneously. Although intuitive, naively applying the student-teacher pipeline does not work well in our task. Not all information from the teacher network is helpful since it is only trained with the old data. To alleviate this problem, we further introduce a random inquirer and an uncertainty-based filter to filter useful information. Our proposed method is general and thus can be adapted to different implicit representations such as neural radiance field (NeRF) and neural surface field. Extensive experimental results for both 3D reconstruction and novel view synthesis demonstrate the effectiveness of our approach compared to different baselines.
Authors: Sara Sabour, Suhani Vora, Daniel Duckworth, Ivan Krasin, David J. Fleet, Andrea Tagliasacchi
Abstract: Neural radiance fields (NeRF) excel at synthesizing new views given multi-view, calibrated images of a static scene. When scenes include distractors, which are not persistent during image capture (moving objects, lighting variations, shadows), artifacts appear as view-dependent effects or 'floaters'. To cope with distractors, we advocate a form of robust estimation for NeRF training, modeling distractors in training data as outliers of an optimization problem. Our method successfully removes outliers from a scene and improves upon our baselines, on synthetic and real-world scenes. Our technique is simple to incorporate in modern NeRF frameworks, with few hyper-parameters. It does not assume a priori knowledge of the types of distractors, and is instead focused on the optimization problem rather than pre-processing or modeling transient objects. More results on our page https://robustnerf.github.io.
Authors: Xiaoyu Zhu, Po-Yao Huang, Junwei Liang, Celso M. de Melo, Alexander Hauptmann
Abstract: We study the problem of human action recognition using motion capture (MoCap) sequences. Unlike existing techniques that take multiple manual steps to derive standardized skeleton representations as model input, we propose a novel Spatial-Temporal Mesh Transformer (STMT) to directly model the mesh sequences. The model uses a hierarchical transformer with intra-frame off-set attention and inter-frame self-attention. The attention mechanism allows the model to freely attend between any two vertex patches to learn non-local relationships in the spatial-temporal domain. Masked vertex modeling and future frame prediction are used as two self-supervised tasks to fully activate the bi-directional and auto-regressive attention in our hierarchical transformer. The proposed method achieves state-of-the-art performance compared to skeleton-based and point-cloud-based models on common MoCap benchmarks. Code is available at https://github.com/zgzxy001/STMT.
Authors: Kai Han, Xiaohu Huang, Yandong Li, Sagar Vaze, Jie Li, Xuhui Jia
Abstract: Existing machine learning models demonstrate excellent performance in image object recognition after training on a large-scale dataset under full supervision. However, these models only learn to map an image to a predefined class index, without revealing the actual semantic meaning of the object in the image. In contrast, vision-language models like CLIP are able to assign semantic class names to unseen objects in a 'zero-shot' manner, though they are once again provided a pre-defined set of candidate names at test-time. In this paper, we reconsider the recognition problem and task a vision-language model with assigning class names to images given only a large (essentially unconstrained) vocabulary of categories as prior information. We leverage non-parametric methods to establish meaningful relationships between images, allowing the model to automatically narrow down the pool of candidate names. Our proposed approach entails iteratively clustering the data and employing a voting mechanism to determine the most suitable class names. Additionally, we investigate the potential of incorporating additional textual features to enhance clustering performance. To achieve this, we employ the CLIP vision and text encoders to retrieve relevant texts from an external database, which can provide supplementary semantic information to inform the clustering process. Furthermore, we tackle this problem both in unsupervised and partially supervised settings, as well as with a coarse-grained and fine-grained search space as the unconstrained dictionary. Remarkably, our method leads to a roughly 50% improvement over the baseline on ImageNet in the unsupervised setting.
Authors: Zehao Zhu, Jiashun Wang, Yuzhe Qin, Deqing Sun, Varun Jampani, Xiaolong Wang
Abstract: We propose a new dataset and a novel approach to learning hand-object interaction priors for hand and articulated object pose estimation. We first collect a dataset using visual teleoperation, where the human operator can directly play within a physical simulator to manipulate the articulated objects. We record the data and obtain free and accurate annotations on object poses and contact information from the simulator. Our system only requires an iPhone to record human hand motion, which can be easily scaled up and largely lower the costs of data and annotation collection. With this data, we learn 3D interaction priors including a discriminator (in a GAN) capturing the distribution of how object parts are arranged, and a diffusion model which generates the contact regions on articulated objects, guiding the hand pose estimation. Such structural and contact priors can easily transfer to real-world data with barely any domain gap. By using our data and learned priors, our method significantly improves the performance on joint hand and articulated object poses estimation over the existing state-of-the-art methods. The project is available at https://zehaozhu.github.io/ContactArt/ .
Authors: Asifullah Khan, Zunaira Rauf, Anabia Sohail, Abdul Rehman, Hifsa Asif, Aqsa Asif, Umair Farooq
Abstract: Vision transformers have become popular as a possible substitute to convolutional neural networks (CNNs) for a variety of computer vision applications. These transformers, with their ability to focus on global relationships in images, offer large learning capacity. However, they may suffer from limited generalization as they do not tend to model local correlation in images. Recently, in vision transformers hybridization of both the convolution operation and self-attention mechanism has emerged, to exploit both the local and global image representations. These hybrid vision transformers, also referred to as CNN-Transformer architectures, have demonstrated remarkable results in vision applications. Given the rapidly growing number of hybrid vision transformers, it has become necessary to provide a taxonomy and explanation of these hybrid architectures. This survey presents a taxonomy of the recent vision transformer architectures and more specifically that of the hybrid vision transformers. Additionally, the key features of these architectures such as the attention mechanisms, positional embeddings, multi-scale processing, and convolution are also discussed. In contrast to the previous survey papers that are primarily focused on individual vision transformer architectures or CNNs, this survey uniquely emphasizes the emerging trend of hybrid vision transformers. By showcasing the potential of hybrid vision transformers to deliver exceptional performance across a range of computer vision tasks, this survey sheds light on the future directions of this rapidly evolving architecture.
Authors: Yifei Ming, Yixuan Li
Abstract: Recent large vision-language models such as CLIP have shown remarkable out-of-distribution (OOD) detection and generalization performance. However, their zero-shot in-distribution (ID) accuracy is often limited for downstream datasets. Recent CLIP-based fine-tuning methods such as prompt learning have demonstrated significant improvements in ID classification and OOD generalization where OOD labels are available. Nonetheless, it remains unclear whether the model is reliable to semantic shifts without OOD labels. In this paper, we aim to bridge the gap and present a comprehensive study to understand how fine-tuning impact OOD detection for few-shot downstream tasks. By framing OOD detection as multi-modal concept matching, we establish a connection between fine-tuning methods and various OOD scores. Our results suggest that a proper choice of OOD scores is essential for CLIP-based fine-tuning. In particular, the maximum concept matching (MCM) score provides a promising solution consistently. We also show that prompt learning demonstrates the state-of-the-art OOD detection performance over the zero-shot counterpart.
Authors: An Dinh Vuong, Toan Tien Nguyen, Minh Nhat VU, Baoru Huang, Dzung Nguyen, Huynh Thi Thanh Binh, Thieu Vo, Anh Nguyen
Abstract: Visual navigation, a foundational aspect of Embodied AI (E-AI), has been significantly studied in the past few years. While many 3D simulators have been introduced to support visual navigation tasks, scarcely works have been directed towards combining human dynamics, creating the gap between simulation and real-world applications. Furthermore, current 3D simulators incorporating human dynamics have several limitations, particularly in terms of computational efficiency, which is a promise of E-AI simulators. To overcome these shortcomings, we introduce HabiCrowd, the first standard benchmark for crowd-aware visual navigation that integrates a crowd dynamics model with diverse human settings into photorealistic environments. Empirical evaluations demonstrate that our proposed human dynamics model achieves state-of-the-art performance in collision avoidance, while exhibiting superior computational efficiency compared to its counterparts. We leverage HabiCrowd to conduct several comprehensive studies on crowd-aware visual navigation tasks and human-robot interactions. The source code and data can be found at https://habicrowd.github.io/.
Authors: Nadav Schneider, Muriel Tzdaka, Galit Sturm, Guy Lazovski, Galit Bar, Gilad Oren, Raz Gvishi, Gal Oren
Abstract: Physical experiments often involve multiple imaging representations, such as X-ray scans and microscopic images. Deep learning models have been widely used for supervised analysis in these experiments. Combining different image representations is frequently required to analyze and make a decision properly. Consequently, multi-view data has emerged - datasets where each sample is described by views from different angles, sources, or modalities. These problems are addressed with the concept of multi-view learning. Understanding the decision-making process of deep learning models is essential for reliable and credible analysis. Hence, many explainability methods have been devised recently. Nonetheless, there is a lack of proper explainability in multi-view models, which are challenging to explain due to their architectures. In this paper, we suggest different multi-view architectures for the vision domain, each suited to another problem, and we also present a methodology for explaining these models. To demonstrate the effectiveness of our methodology, we focus on the domain of High Energy Density Physics (HEDP) experiments, where multiple imaging representations are used to assess the quality of foam samples. We apply our methodology to classify the foam samples quality using the suggested multi-view architectures. Through experimental results, we showcase the improvement of accurate architecture choice on both accuracy - 78% to 84% and AUC - 83% to 93% and present a trade-off between performance and explainability. Specifically, we demonstrate that our approach enables the explanation of individual one-view models, providing insights into the decision-making process of each view. This understanding enhances the interpretability of the overall multi-view model. The sources of this work are available at: https://github.com/Scientific-Computing-Lab-NRCN/Multi-View-Explainability.
URLs: https://github.com/Scientific-Computing-Lab-NRCN/Multi-View-Explainability.
Authors: Giuseppe Vecchio, Rosalie Martin, Arthur Roullier, Adrien Kaiser, Romain Rouffet, Valentin Deschaintre, Tamy Boubekeur
Abstract: Material reconstruction from a photograph is a key component of 3D content creation democratization. We propose to formulate this ill-posed problem as a controlled synthesis one, leveraging the recent progress in generative deep networks. We present ControlMat, a method which, given a single photograph with uncontrolled illumination as input, conditions a diffusion model to generate plausible, tileable, high-resolution physically-based digital materials. We carefully analyze the behavior of diffusion models for multi-channel outputs, adapt the sampling process to fuse multi-scale information and introduce rolled diffusion to enable both tileability and patched diffusion for high-resolution outputs. Our generative approach further permits exploration of a variety of materials which could correspond to the input image, mitigating the unknown lighting conditions. We show that our approach outperforms recent inference and latent-space-optimization methods, and carefully validate our diffusion process design choices. Supplemental materials and additional details are available at: https://gvecchio.com/controlmat/.
Authors: Shilin Yan, Xiaohao Xu, Renrui Zhang, Lingyi Hong, Wenchao Chen, Wenqiang Zhang, Wei Zhang
Abstract: Panoramic videos contain richer spatial information and have attracted tremendous amounts of attention due to their exceptional experience in some fields such as autonomous driving and virtual reality. However, existing datasets for video segmentation only focus on conventional planar images. To address the challenge, in this paper, we present a panoramic video dataset, PanoVOS. The dataset provides 150 videos with high video resolutions and diverse motions. To quantify the domain gap between 2D planar videos and panoramic videos, we evaluate 15 off-the-shelf video object segmentation (VOS) models on PanoVOS. Through error analysis, we found that all of them fail to tackle pixel-level content discontinues of panoramic videos. Thus, we present a Panoramic Space Consistency Transformer (PSCFormer), which can effectively utilize the semantic boundary information of the previous frame for pixel-level matching with the current frame. Extensive experiments demonstrate that compared with the previous SOTA models, our PSCFormer network exhibits a great advantage in terms of segmentation results under the panoramic setting. Our dataset poses new challenges in panoramic VOS and we hope that our PanoVOS can advance the development of panoramic segmentation/tracking.
Authors: Shaocong Long, Qianyu Zhou, Chenhao Ying, Lizhuang Ma, Yuan Luo
Abstract: Domain generalization(DG) endeavors to develop robust models that possess strong generalizability while preserving excellent discriminability. Nonetheless, pivotal DG techniques tend to improve the feature generalizability by learning domain-invariant representations, inadvertently overlooking the feature discriminability. On the one hand, the simultaneous attainment of generalizability and discriminability of features presents a complex challenge, often entailing inherent contradictions. This challenge becomes particularly pronounced when domain-invariant features manifest reduced discriminability owing to the inclusion of unstable factors, i.e., spurious correlations. On the other hand, prevailing domain-invariant methods can be categorized as category-level alignment, susceptible to discarding indispensable features possessing substantial generalizability and narrowing intra-class variations. To surmount these obstacles, we rethink DG from a new perspective that concurrently imbues features with formidable discriminability and robust generalizability, and present a novel framework, namely, Discriminative Microscopic Distribution Alignment~(DMDA). DMDA incorporates two core components: Selective Channel Pruning~(SCP) and Micro-level Distribution Alignment~(MDA). Concretely, SCP attempts to curtail redundancy within neural networks, prioritizing stable attributes conducive to accurate classification. This approach alleviates the adverse effect of spurious domain invariance and amplifies the feature discriminability. Besides, MDA accentuates micro-level alignment within each class, going beyond mere category-level alignment. Extensive experiments on four benchmark datasets corroborate that DMDA achieves comparable results to state-of-the-art methods in DG, underscoring the efficacy of our method.
Authors: Nakyeong Yang, Minsung Kim, Seunghyun Yoon, Joongbo Shin, Kyomin Jung
Abstract: With the explosion of multimedia content, video moment retrieval (VMR), which aims to detect a video moment that matches a given text query from a video, has been studied intensively as a critical problem. However, the existing VMR framework evaluates video moment retrieval performance, assuming that a video is given, which may not reveal whether the models exhibit overconfidence in the falsely given video. In this paper, we propose the MVMR (Massive Videos Moment Retrieval for Faithfulness Evaluation) task that aims to retrieve video moments within a massive video set, including multiple distractors, to evaluate the faithfulness of VMR models. For this task, we suggest an automated massive video pool construction framework to categorize negative (distractors) and positive (false-negative) video sets using textual and visual semantic distance verification methods. We extend existing VMR datasets using these methods and newly construct three practical MVMR datasets. To solve the task, we further propose a strong informative sample-weighted learning method, CroCs, which employs two contrastive learning mechanisms: (1) weakly-supervised potential negative learning and (2) cross-directional hard-negative learning. Experimental results on the MVMR datasets reveal that existing VMR models are easily distracted by the misinformation (distractors), whereas our model shows significantly robust performance, demonstrating that CroCs is essential to distinguishing positive moments against distractors. Our code and datasets are publicly available: https://github.com/yny0506/Massive-Videos-Moment-Retrieval.
URLs: https://github.com/yny0506/Massive-Videos-Moment-Retrieval.
Authors: Zhongang Cai, Wanqi Yin, Ailing Zeng, Chen Wei, Qingping Sun, Yanjun Wang, Hui En Pang, Haiyi Mei, Mingyuan Zhang, Lei Zhang, Chen Change Loy, Lei Yang, Ziwei Liu
Abstract: Expressive human pose and shape estimation (EHPS) unifies body, hands, and face motion capture with numerous applications. Despite encouraging progress, current state-of-the-art methods still depend largely on a confined set of training datasets. In this work, we investigate scaling up EHPS towards the first generalist foundation model (dubbed SMPLer-X), with up to ViT-Huge as the backbone and training with up to 4.5M instances from diverse data sources. With big data and the large model, SMPLer-X exhibits strong performance across diverse test benchmarks and excellent transferability to even unseen environments. 1) For the data scaling, we perform a systematic investigation on 32 EHPS datasets, including a wide range of scenarios that a model trained on any single dataset cannot handle. More importantly, capitalizing on insights obtained from the extensive benchmarking process, we optimize our training scheme and select datasets that lead to a significant leap in EHPS capabilities. 2) For the model scaling, we take advantage of vision transformers to study the scaling law of model sizes in EHPS. Moreover, our finetuning strategy turn SMPLer-X into specialist models, allowing them to achieve further performance boosts. Notably, our foundation model SMPLer-X consistently delivers state-of-the-art results on seven benchmarks such as AGORA (107.2 mm NMVE), UBody (57.4 mm PVE), EgoBody (63.6 mm PVE), and EHF (62.3 mm PVE without finetuning). Homepage: https://caizhongang.github.io/projects/SMPLer-X/
Authors: Md Kaykobad Reza, Ashley Prater-Bennette, M. Salman Asif
Abstract: Multimodal learning seeks to utilize data from multiple sources to improve the overall performance of downstream tasks. It is desirable for redundancies in the data to make multimodal systems robust to missing or corrupted observations in some correlated modalities. However, we observe that the performance of several existing multimodal networks significantly deteriorates if one or multiple modalities are absent at test time. To enable robustness to missing modalities, we propose a simple and parameter-efficient adaptation procedure for pretrained multimodal networks. In particular, we exploit modulation of intermediate features to compensate for the missing modalities. We demonstrate that such adaptation can partially bridge performance drop due to missing modalities and outperform independent, dedicated networks trained for the available modality combinations in some cases. The proposed adaptation requires extremely small number of parameters (e.g., fewer than 1% of the total parameters) and applicable to a wide range of modality combinations and tasks. We conduct a series of experiments to highlight the missing modality robustness of our proposed method on five different multimodal tasks across seven datasets. Our proposed method demonstrates versatility across various tasks and datasets, and outperforms existing methods for robust multimodal learning with missing modalities.
Authors: Jiageng Mao, Junjie Ye, Yuxi Qian, Marco Pavone, Yue Wang
Abstract: Human-level driving is an ultimate goal of autonomous driving. Conventional approaches formulate autonomous driving as a perception-prediction-planning framework, yet their systems do not capitalize on the inherent reasoning ability and experiential knowledge of humans. In this paper, we propose a fundamental paradigm shift from current pipelines, exploiting Large Language Models (LLMs) as a cognitive agent to integrate human-like intelligence into autonomous driving systems. Our approach, termed Agent-Driver, transforms the traditional autonomous driving pipeline by introducing a versatile tool library accessible via function calls, a cognitive memory of common sense and experiential knowledge for decision-making, and a reasoning engine capable of chain-of-thought reasoning, task planning, motion planning, and self-reflection. Powered by LLMs, our Agent-Driver is endowed with intuitive common sense and robust reasoning capabilities, thus enabling a more nuanced, human-like approach to autonomous driving. We evaluate our approach on the large-scale nuScenes benchmark, and extensive experiments substantiate that our Agent-Driver significantly outperforms the state-of-the-art driving methods by a large margin. Our approach also demonstrates superior interpretability and few-shot learning ability to these methods.
Authors: Zhixi Cai, Shreya Ghosh, Aman Pankaj Adatia, Munawar Hayat, Abhinav Dhall, Tom Gedeon, Kalin Stefanov
Abstract: The detection and localization of highly realistic deepfake audio-visual content are challenging even for the most advanced state-of-the-art methods. While most of the research efforts in this domain are focused on detecting high-quality deepfake images and videos, only a few works address the problem of the localization of small segments of audio-visual manipulations embedded in real videos. In this research, we emulate the process of such content generation and propose the AV-Deepfake1M dataset. The dataset contains content-driven (i) video manipulations, (ii) audio manipulations, and (iii) audio-visual manipulations for more than 2K subjects resulting in a total of more than 1M videos. The paper provides a thorough description of the proposed data generation pipeline accompanied by a rigorous analysis of the quality of the generated data. The comprehensive benchmark of the proposed dataset utilizing state-of-the-art deepfake detection and localization methods indicates a significant drop in performance compared to previous datasets. The proposed dataset will play a vital role in building the next-generation deepfake localization methods. The dataset and associated code are available at https://github.com/ControlNet/AV-Deepfake1M .
Authors: Yang Liu, Xiang Huang, Minghan Qin, Qinwei Lin, Haoqian Wang
Abstract: Neural radiance fields are capable of reconstructing high-quality drivable human avatars but are expensive to train and render and not suitable for multi-human scenes with complex shadows. To reduce consumption, we propose Animatable 3D Gaussian, which learns human avatars from input images and poses. We extend 3D Gaussians to dynamic human scenes by modeling a set of skinned 3D Gaussians and a corresponding skeleton in canonical space and deforming 3D Gaussians to posed space according to the input poses. We introduce a multi-head hash encoder for pose-dependent shape and appearance and a time-dependent ambient occlusion module to achieve high-quality reconstructions in scenes containing complex motions and dynamic shadows. On both novel view synthesis and novel pose synthesis tasks, our method achieves higher reconstruction quality than InstantAvatar with less training time (1/60), less GPU memory (1/4), and faster rendering speed (7x). Our method can be easily extended to multi-human scenes and achieve comparable novel view synthesis results on a scene with ten people in only 25 seconds of training.
Authors: Bin Xia, Shiyin Wang, Yingfan Tao, Yitong Wang, Jiaya Jia
Abstract: In this paper, we introduce a Multimodal Large Language Model-based Generation Assistant (LLMGA), leveraging the vast reservoir of knowledge and proficiency in reasoning, comprehension, and response inherent in Large Language Models (LLMs) to assist users in image generation and editing. Diverging from existing approaches where Multimodal Large Language Models (MLLMs) generate fixed-size embeddings to control Stable Diffusion (SD), our LLMGA provides a detailed language generation prompt for precise control over SD. This not only augments LLM context understanding but also reduces noise in generation prompts, yields images with more intricate and precise content, and elevates the interpretability of the network. To this end, we curate a comprehensive dataset comprising prompt refinement, similar image generation, inpainting \& outpainting, and instruction-based editing. Moreover, we propose a two-stage training scheme. In the first stage, we train the MLLM to grasp the properties of image generation and editing, enabling it to generate detailed prompts. In the second stage, we optimize SD to align with the MLLM's generation prompts. Additionally, we propose a reference-based restoration network to alleviate texture, brightness, and contrast disparities between generated and preserved regions during inpainting and outpainting. Extensive results show that LLMGA has promising generation and editing capabilities and can enable more flexible and expansive applications in an interactive manner.
Authors: Anindita Ghosh, Rishabh Dabral, Vladislav Golyanik, Christian Theobalt, Philipp Slusallek
Abstract: Current approaches for 3D human motion synthesis generate high quality animations of digital humans performing a wide variety of actions and gestures. However, a notable technological gap exists in addressing the complex dynamics of multi human interactions within this paradigm. In this work, we present ReMoS, a denoising diffusion based model that synthesizes full body reactive motion of a person in a two person interaction scenario. Given the motion of one person, we employ a combined spatio temporal cross attention mechanism to synthesize the reactive body and hand motion of the second person, thereby completing the interactions between the two. We demonstrate ReMoS across challenging two person scenarios such as pair dancing, Ninjutsu, kickboxing, and acrobatics, where one persons movements have complex and diverse influences on the other. We also contribute the ReMoCap dataset for two person interactions containing full body and finger motions. We evaluate ReMoS through multiple quantitative metrics, qualitative visualizations, and a user study, and also indicate usability in interactive motion editing applications.
Authors: Rao Fu, Zehao Wen, Zichen Liu, Srinath Sridhar
Abstract: Inspired by cognitive theories, we introduce AnyHome, a framework that translates any text into well-structured and textured indoor scenes at a house-scale. By prompting Large Language Models (LLMs) with designed templates, our approach converts provided textual narratives into amodal structured representations. These representations guarantee consistent and realistic spatial layouts by directing the synthesis of a geometry mesh within defined constraints. A Score Distillation Sampling process is then employed to refine the geometry, followed by an egocentric inpainting process that adds lifelike textures to it. AnyHome stands out with its editability, customizability, diversity, and realism. The structured representations for scenes allow for extensive editing at varying levels of granularity. Capable of interpreting texts ranging from simple labels to detailed narratives, AnyHome generates detailed geometries and textures that outperform existing methods in both quantitative and qualitative measures.
Authors: Yuhui Wu, Guoqing Wang, Zhiwen Wang, Yang Yang, Tianyu Li, Malu Zhang, Chongyi Li, Heng Tao Shen
Abstract: Low-light image enhancement (LLIE) has achieved promising performance by employing conditional diffusion models. Despite the success of some conditional methods, previous methods may neglect the importance of a sufficient formulation of task-specific condition strategy, resulting in suboptimal visual outcomes. In this study, we propose JoReS-Diff, a novel approach that incorporates Retinex- and semantic-based priors as the additional pre-processing condition to regulate the generating capabilities of the diffusion model. We first leverage pre-trained decomposition network to generate the Retinex prior, which is updated with better quality by an adjustment network and integrated into a refinement network to implement Retinex-based conditional generation at both feature- and image-levels. Moreover, the semantic prior is extracted from the input image with an off-the-shelf semantic segmentation model and incorporated through semantic attention layers. By treating Retinex- and semantic-based priors as the condition, JoReS-Diff presents a unique perspective for establishing an diffusion model for LLIE and similar image enhancement tasks. Extensive experiments validate the rationality and superiority of our approach.
Authors: Hao Liu, Jun Liu, Raymond H. Chan, Xue-Cheng Tai
Abstract: In this study, our goal is to integrate classical mathematical models with deep neural networks by introducing two novel deep neural network models for image segmentation known as Double-well Nets. Drawing inspirations from the Potts model, our models leverage neural networks to represent a region force functional. We extend the well-know MBO (Merriman-Bence-Osher) scheme to solve the Potts model. The widely recognized Potts model is approximated using a double-well potential and then solved by an operator-splitting method, which turns out to be an extension of the well-known MBO scheme. Subsequently, we replace the region force functional in the Potts model with a UNet-type network, which is data-driven and is designed to capture multiscale features of images, and also introduce control variables to enhance effectiveness. The resulting algorithm is a neural network activated by a function that minimizes the double-well potential. What sets our proposed Double-well Nets apart from many existing deep learning methods for image segmentation is their strong mathematical foundation. They are derived from the network approximation theory and employ the MBO scheme to approximately solve the Potts model. By incorporating mathematical principles, Double-well Nets bridge the MBO scheme and neural networks, and offer an alternative perspective for designing networks with mathematical backgrounds. Through comprehensive experiments, we demonstrate the performance of Double-well Nets, showcasing their superior accuracy and robustness compared to state-of-the-art neural networks. Overall, our work represents a valuable contribution to the field of image segmentation by combining the strengths of classical variational models and deep neural networks. The Double-well Nets introduce an innovative approach that leverages mathematical foundations to enhance segmentation performance.
Authors: Jing Wu, Trung Le, Munawar Hayat, Mehrtash Harandi
Abstract: We introduce EraseDiff, an unlearning algorithm designed for diffusion models to address concerns related to data memorization. Our approach formulates the unlearning task as a constrained optimization problem, aiming to preserve the utility of the diffusion model on retained data while removing the information associated with the data to be forgotten. This is achieved by altering the generative process to deviate away from the ground-truth denoising procedure. To manage the computational complexity inherent in the diffusion process, we develop a first-order method for solving the optimization problem, which has shown empirical benefits. Extensive experiments and thorough comparisons with state-of-the-art algorithms demonstrate that EraseDiff effectively preserves the model's utility, efficacy, and efficiency.
Authors: Bin Dou, Tianyu Zhang, Zhaohui Wang, Yongjia Ma, Zejian Yuan
Abstract: Zero-shot neural scene segmentation, which reconstructs 3D neural segmentation field without manual annotations, serves as an effective way for scene understanding. However, existing models, especially the efficient 3D Gaussian-based methods, struggle to produce compact segmentation results. This issue stems primarily from their redundant learnable attributes assigned on individual Gaussians, leading to a lack of robustness against the 3D-inconsistencies in zero-shot generated raw labels. To address this problem, our work, named Compact Segmented 3D Gaussians (CoSegGaussians), proposes the Feature Unprojection and Fusion module as the segmentation field, which utilizes a shallow decoder generalizable for all Gaussians based on high-level features. Specifically, leveraging the learned Gaussian geometric parameters, semantic-aware image-based features are introduced into the scene via our unprojection technique. The lifted features, together with spatial information, are fed into the multi-scale aggregation decoder to generate segmentation identities for all Gaussians. Furthermore, we design CoSeg Loss to boost model robustness against 3D-inconsistent noises. Experimental results show that our model surpasses baselines on zero-shot semantic segmentation task, improving by ~10% mIoU over the best baseline. Code and more results will be available at https://David-Dou.github.io/CoSegGaussians.
Authors: Jiangming Shi, Xiangbo Yin, Yeyun Chen, Yachao Zhang, Zhizhong Zhang, Yuan Xie, Yanyun Qu
Abstract: Unsupervised visible-infrared person re-identification (USL-VI-ReID) is a promising yet challenging retrieval task. The key challenges in USL-VI-ReID are to effectively generate pseudo-labels and establish pseudo-label correspondences across modalities without relying on any prior annotations. Recently, clustered pseudo-label methods have gained more attention in USL-VI-ReID. However, previous methods fell short of fully exploiting the individual nuances, as they simply utilized a single memory that represented an identity to establish cross-modality correspondences, resulting in ambiguous cross-modality correspondences. To address the problem, we propose a Multi-Memory Matching (MMM) framework for USL-VI-ReID. We first design a Cross-Modality Clustering (CMC) module to generate the pseudo-labels through clustering together both two modality samples. To associate cross-modality clustered pseudo-labels, we design a Multi-Memory Learning and Matching (MMLM) module, ensuring that optimization explicitly focuses on the nuances of individual perspectives and establishes reliable cross-modality correspondences. Finally, we design a Soft Cluster-level Alignment (SCA) module to narrow the modality gap while mitigating the effect of noise pseudo-labels through a soft many-to-many alignment strategy. Extensive experiments on the public SYSU-MM01 and RegDB datasets demonstrate the reliability of the established cross-modality correspondences and the effectiveness of our MMM. The source codes will be released.
Authors: Kibum Kim, Kanghoon Yoon, Yeonjun In, Jinyoung Moon, Donghyun Kim, Chanyoung Park
Abstract: Scene graph generation (SGG) models have suffered from inherent problems regarding the benchmark datasets such as the long-tailed predicate distribution and missing annotation problems. In this work, we aim to alleviate the long-tailed problem of SGG by utilizing unannotated triplets. To this end, we introduce a Self-Training framework for SGG (ST-SGG) that assigns pseudo-labels for unannotated triplets based on which the SGG models are trained. While there has been significant progress in self-training for image recognition, designing a self-training framework for the SGG task is more challenging due to its inherent nature such as the semantic ambiguity and the long-tailed distribution of predicate classes. Hence, we propose a novel pseudo-labeling technique for SGG, called Class-specific Adaptive Thresholding with Momentum (CATM), which is a model-agnostic framework that can be applied to any existing SGG models. Furthermore, we devise a graph structure learner (GSL) that is beneficial when adopting our proposed self-training framework to the state-of-the-art message-passing neural network (MPNN)-based SGG models. Our extensive experiments verify the effectiveness of ST-SGG on various SGG models, particularly in enhancing the performance on fine-grained predicate classes.
Authors: Hongchi Xia, Yang Fu, Sifei Liu, Xiaolong Wang
Abstract: We introduce a new RGB-D object dataset captured in the wild called WildRGB-D. Unlike most existing real-world object-centric datasets which only come with RGB capturing, the direct capture of the depth channel allows better 3D annotations and broader downstream applications. WildRGB-D comprises large-scale category-level RGB-D object videos, which are taken using an iPhone to go around the objects in 360 degrees. It contains around 8500 recorded objects and nearly 20000 RGB-D videos across 46 common object categories. These videos are taken with diverse cluttered backgrounds with three setups to cover as many real-world scenarios as possible: (i) a single object in one video; (ii) multiple objects in one video; and (iii) an object with a static hand in one video. The dataset is annotated with object masks, real-world scale camera poses, and reconstructed aggregated point clouds from RGBD videos. We benchmark four tasks with WildRGB-D including novel view synthesis, camera pose estimation, object 6d pose estimation, and object surface reconstruction. Our experiments show that the large-scale capture of RGB-D objects provides a large potential to advance 3D object learning. Our project page is https://wildrgbd.github.io/.
Authors: Yiran Song, Qianyu Zhou, Xuequan Lu, Zhiwen Shao, Lizhuang Ma
Abstract: Segment anything model (SAM) has demonstrated excellent generalizability in common vision scenarios, yet falling short of the ability to understand specialized data. Recently, several methods have combined parameter-efficient techniques with task-specific designs to fine-tune SAM on particular tasks. However, these methods heavily rely on handcraft, complicated, and task-specific designs, and pre/post-processing to achieve acceptable performances on downstream tasks. As a result, this severely restricts generalizability to other downstream tasks. To address this issue, we present a simple and unified framework, namely SU-SAM, that can easily and efficiently fine-tune the SAM model with parameter-efficient techniques while maintaining excellent generalizability toward various downstream tasks. SU-SAM does not require any task-specific designs and aims to improve the adaptability of SAM-like models significantly toward underperformed scenes. Concretely, we abstract parameter-efficient modules of different methods into basic design elements in our framework. Besides, we propose four variants of SU-SAM, i.e., series, parallel, mixed, and LoRA structures. Comprehensive experiments on nine datasets and six downstream tasks to verify the effectiveness of SU-SAM, including medical image segmentation, camouflage object detection, salient object segmentation, surface defect segmentation, complex object shapes, and shadow masking. Our experimental results demonstrate that SU-SAM achieves competitive or superior accuracy compared to state-of-the-art methods. Furthermore, we provide in-depth analyses highlighting the effectiveness of different parameter-efficient designs within SU-SAM. In addition, we propose a generalized model and benchmark, showcasing SU-SAM's generalizability across all diverse datasets simultaneously.
Authors: Yang Liu, Xinshuai Song, Kaixuan Jiang, Weixing Chen, Jingzhou Luo, Guanbin Li, Liang Lin
Abstract: With the surge in the development of large language models, embodied intelligence has attracted increasing attention. Nevertheless, prior works on embodied intelligence typically encode scene or historical memory in an unimodal manner, either visual or linguistic, which complicates the alignment of the model's action planning with embodied control. To overcome this limitation, we introduce the Multimodal Embodied Interactive Agent (MEIA), capable of translating high-level tasks expressed in natural language into a sequence of executable actions. Specifically, we propose a novel Multimodal Environment Memory (MEM) module, facilitating the integration of embodied control with large models through the visual-language memory of scenes. This capability enables MEIA to generate executable action plans based on diverse requirements and the robot's capabilities. Furthermore, we construct an embodied question answering dataset based on a dynamic virtual cafe environment with the help of the large language model. In this virtual environment, we conduct several experiments, utilizing multiple large models through zero-shot learning, and carefully design scenarios for various situations. The experimental results showcase the promising performance of our MEIA in various embodied interactive tasks.
Authors: Weiliang Chen, Qianqian Ren, Lin Pan, Shengxi Fu, Jinbao Li
Abstract: Urban region profiling is influential for smart cities and sustainable development. However, extracting fine-grained semantics and generating robust urban region embeddings from noisy and incomplete urban data is challenging. In response, we present EUPAC (Enhanced Urban Region Profiling with Adversarial Contrastive Learning), a novel framework that enhances the robustness of urban region embeddings through joint optimization of attentive supervised and adversarial contrastive modules. Specifically, region heterogeneous graphs containing human mobility data, point of interest information, and geographic neighborhood details for each region are fed into our model, which generates region embeddings that preserve intra-region and inter-region dependencies through graph convolutional networks and multi-head attention. Meanwhile, we introduce spatially learnable augmentation to generate positive samples that are semantically similar and spatially close to the anchor, preparing for subsequent contrastive learning. Furthermore, we propose an adversarial training method to construct an effective pretext task by generating strong positive pairs and mining hard negative pairs for the region embeddings. Finally, we jointly optimize attentive supervised and adversarial contrastive learning to encourage the model to capture the high-level semantics of region embeddings while ignoring the noisy and irrelevant details. Extensive experiments on real-world datasets demonstrate the superiority of our model over state-of-the-art methods.
Authors: Zhicheng Zheng, Xin Yan, Zhenfang Chen, Jingzhou Wang, Qin Zhi Eddie Lim, Joshua B. Tenenbaum, Chuang Gan
Abstract: We introduce the Continuum Physical Dataset (ContPhy), a novel benchmark for assessing machine physical commonsense. ContPhy complements existing physical reasoning benchmarks by encompassing the inference of diverse physical properties, such as mass and density, across various scenarios and predicting corresponding dynamics. We evaluated a range of AI models and found that they still struggle to achieve satisfactory performance on ContPhy, which shows that the current AI models still lack physical commonsense for the continuum, especially soft-bodies, and illustrates the value of the proposed dataset. We also introduce an oracle model (ContPRO) that marries the particle-based physical dynamic models with the recent large language models, which enjoy the advantages of both models, precise dynamic predictions, and interpretable reasoning. ContPhy aims to spur progress in perception and reasoning within diverse physical settings, narrowing the divide between human and machine intelligence in understanding the physical world. Project page: https://physical-reasoning-project.github.io
Authors: Chun-Lin Ji, Tao Yu, Peng Gao, Fei Wang, Ru-Yue Yuan
Abstract: Object detection, a crucial aspect of computer vision, has seen significant advancements in accuracy and robustness. Despite these advancements, practical applications still face notable challenges, primarily the inaccurate detection or missed detection of small objects. In this paper, we propose YOLO-TLA, an advanced object detection model building on YOLOv5. We first introduce an additional detection layer for small objects in the neck network pyramid architecture, thereby producing a feature map of a larger scale to discern finer features of small objects. Further, we integrate the C3CrossCovn module into the backbone network. This module uses sliding window feature extraction, which effectively minimizes both computational demand and the number of parameters, rendering the model more compact. Additionally, we have incorporated a global attention mechanism into the backbone network. This mechanism combines the channel information with global information to create a weighted feature map. This feature map is tailored to highlight the attributes of the object of interest, while effectively ignoring irrelevant details. In comparison to the baseline YOLOv5s model, our newly developed YOLO-TLA model has shown considerable improvements on the MS COCO validation dataset, with increases of 4.6% in mAP@0.5 and 4% in mAP@0.5:0.95, all while keeping the model size compact at 9.49M parameters. Further extending these improvements to the YOLOv5m model, the enhanced version exhibited a 1.7% and 1.9% increase in mAP@0.5 and mAP@0.5:0.95, respectively, with a total of 27.53M parameters. These results validate the YOLO-TLA model's efficient and effective performance in small object detection, achieving high accuracy with fewer parameters and computational demands.
Authors: Lukas H\"ollein, Alja\v{z} Bo\v{z}i\v{c}, Norman M\"uller, David Novotny, Hung-Yu Tseng, Christian Richardt, Michael Zollh\"ofer, Matthias Nie{\ss}ner
Abstract: 3D asset generation is getting massive amounts of attention, inspired by the recent success of text-guided 2D content creation. Existing text-to-3D methods use pretrained text-to-image diffusion models in an optimization problem or fine-tune them on synthetic data, which often results in non-photorealistic 3D objects without backgrounds. In this paper, we present a method that leverages pretrained text-to-image models as a prior, and learn to generate multi-view images in a single denoising process from real-world data. Concretely, we propose to integrate 3D volume-rendering and cross-frame-attention layers into each block of the existing U-Net network of the text-to-image model. Moreover, we design an autoregressive generation that renders more 3D-consistent images at any viewpoint. We train our model on real-world datasets of objects and showcase its capabilities to generate instances with a variety of high-quality shapes and textures in authentic surroundings. Compared to the existing methods, the results generated by our method are consistent, and have favorable visual quality (-30% FID, -37% KID).
Authors: Yujia Liu, Anton Obukhov, Jan Dirk Wegner, Konrad Schindler
Abstract: We present a learning-based approach to reconstruct buildings as 3D polygonal meshes from airborne LiDAR point clouds. What makes 3D building reconstruction from airborne LiDAR hard is the large diversity of building designs and especially roof shapes, the low and varying point density across the scene, and the often incomplete coverage of building facades due to occlusions by vegetation or to the viewing angle of the sensor. To cope with the diversity of shapes and inhomogeneous and incomplete object coverage, we introduce a generative model that directly predicts 3D polygonal meshes from input point clouds. Our autoregressive model, called Point2Building, iteratively builds up the mesh by generating sequences of vertices and faces. This approach enables our model to adapt flexibly to diverse geometries and building structures. Unlike many existing methods that rely heavily on pre-processing steps like exhaustive plane detection, our model learns directly from the point cloud data, thereby reducing error propagation and increasing the fidelity of the reconstruction. We experimentally validate our method on a collection of airborne LiDAR data of Zurich, Berlin and Tallinn. Our method shows good generalization to diverse urban styles.
Authors: Yunhao Li, Qin Li, Hao Wang, Xue Ma, Jiali Yao, Shaohua Dong, Heng Fan, Libo Zhang
Abstract: Current multi-object tracking (MOT) aims to predict trajectories of targets (i.e., ''where'') in videos. Yet, knowing merely ''where'' is insufficient in many crucial applications. In comparison, semantic understanding such as fine-grained behaviors, interactions, and overall summarized captions (i.e., ''what'') from videos, associated with ''where'', is highly-desired for comprehensive video analysis. Thus motivated, we introduce Semantic Multi-Object Tracking (SMOT), that aims to estimate object trajectories and meanwhile understand semantic details of associated trajectories including instance captions, instance interactions, and overall video captions, integrating ''where'' and ''what'' for tracking. In order to foster the exploration of SMOT, we propose BenSMOT, a large-scale Benchmark for Semantic MOT. Specifically, BenSMOT comprises 3,292 videos with 151K frames, covering various scenarios for semantic tracking of humans. BenSMOT provides annotations for the trajectories of targets, along with associated instance captions in natural language, instance interactions, and overall caption for each video sequence. To our best knowledge, BenSMOT is the first publicly available benchmark for SMOT. Besides, to encourage future research, we present a novel tracker named SMOTer, which is specially designed and end-to-end trained for SMOT, showing promising performance. By releasing BenSMOT, we expect to go beyond conventional MOT by predicting ''where'' and ''what'' for SMOT, opening up a new direction in tracking for video understanding. We will release BenSMOT and SMOTer at https://github.com/Nathan-Li123/SMOTer.
Authors: Yisol Choi, Sangkyung Kwak, Kyungmin Lee, Hyungwon Choi, Jinwoo Shin
Abstract: This paper considers image-based virtual try-on, which renders an image of a person wearing a curated garment, given a pair of images depicting the person and the garment, respectively. Previous works adapt existing exemplar-based inpainting diffusion models for virtual try-on to improve the naturalness of the generated visuals compared to other methods (e.g., GAN-based), but they fail to preserve the identity of the garments. To overcome this limitation, we propose a novel diffusion model that improves garment fidelity and generates authentic virtual try-on images. Our method, coined IDM-VTON, uses two different modules to encode the semantics of garment image; given the base UNet of the diffusion model, 1) the high-level semantics extracted from a visual encoder are fused to the cross-attention layer, and then 2) the low-level features extracted from parallel UNet are fused to the self-attention layer. In addition, we provide detailed textual prompts for both garment and person images to enhance the authenticity of the generated visuals. Finally, we present a customization method using a pair of person-garment images, which significantly improves fidelity and authenticity. Our experimental results show that our method outperforms previous approaches (both diffusion-based and GAN-based) in preserving garment details and generating authentic virtual try-on images, both qualitatively and quantitatively. Furthermore, the proposed customization method demonstrates its effectiveness in a real-world scenario. More visualizations are available in our project page: https://idm-vton.github.io
Authors: Yash Bhalgat, Iro Laina, Jo\~ao F. Henriques, Andrew Zisserman, Andrea Vedaldi
Abstract: Understanding complex scenes at multiple levels of abstraction remains a formidable challenge in computer vision. To address this, we introduce Nested Neural Feature Fields (N2F2), a novel approach that employs hierarchical supervision to learn a single feature field, wherein different dimensions within the same high-dimensional feature encode scene properties at varying granularities. Our method allows for a flexible definition of hierarchies, tailored to either the physical dimensions or semantics or both, thereby enabling a comprehensive and nuanced understanding of scenes. We leverage a 2D class-agnostic segmentation model to provide semantically meaningful pixel groupings at arbitrary scales in the image space, and query the CLIP vision-encoder to obtain language-aligned embeddings for each of these segments. Our proposed hierarchical supervision method then assigns different nested dimensions of the feature field to distill the CLIP embeddings using deferred volumetric rendering at varying physical scales, creating a coarse-to-fine representation. Extensive experiments show that our approach outperforms the state-of-the-art feature field distillation methods on tasks such as open-vocabulary 3D segmentation and localization, demonstrating the effectiveness of the learned nested feature field.
Authors: Lachlan Thorpe, Lewis Bawden, Karanjot Vendal, John Bronskill, Richard E. Turner
Abstract: We present a transformer decoder based sports simulation engine, SportsNGEN, trained on sports player and ball tracking sequences, that is capable of generating sustained gameplay and accurately mimicking the decision making of real players. By training on a large database of professional tennis tracking data, we demonstrate that simulations produced by SportsNGEN can be used to predict the outcomes of rallies, determine the best shot choices at any point, and evaluate counterfactual or what if scenarios to inform coaching decisions and elevate broadcast coverage. By combining the generated simulations with a shot classifier and logic to start and end rallies, the system is capable of simulating an entire tennis match. We evaluate SportsNGEN by comparing statistics of the simulations with those of real matches between the same players. We show that the model output sampling parameters are crucial to simulation realism and that SportsNGEN is probabilistically well-calibrated to real data. In addition, a generic version of SportsNGEN can be customized to a specific player by fine-tuning on the subset of match data that includes that player. Finally, we show qualitative results indicating the same approach works for football.
Authors: Wenjun Huang, Hanning Chen, Yang Ni, Arghavan Rezvani, Sanggeon Yun, Sungheon Jeon, Eric Pedley, Mohsen Imani
Abstract: Detecting marine objects inshore presents challenges owing to algorithmic intricacies and complexities in system deployment. We propose a difficulty-aware edge-cloud collaborative sensing system that splits the task into object localization and fine-grained classification. Objects are classified either at the edge or within the cloud, based on their estimated difficulty. The framework comprises a low-power device-tailored front-end model for object localization, classification, and difficulty estimation, along with a transformer-graph convolutional network-based back-end model for fine-grained classification. Our system demonstrates superior performance (mAP@0.5 +4.3%}) on widely used marine object detection datasets, significantly reducing both data transmission volume (by 95.43%) and energy consumption (by 72.7%}) at the system level. We validate the proposed system across various embedded system platforms and in real-world scenarios involving drone deployment.
Authors: Yi Wang, Kunchang Li, Xinhao Li, Jiashuo Yu, Yinan He, Chenting Wang, Guo Chen, Baoqi Pei, Ziang Yan, Rongkun Zheng, Jilan Xu, Zun Wang, Yansong Shi, Tianxiang Jiang, Songze Li, Hongjie Zhang, Yifei Huang, Yu Qiao, Yali Wang, Limin Wang
Abstract: We introduce InternVideo2, a new family of video foundation models (ViFM) that achieve the state-of-the-art results in video recognition, video-text tasks, and video-centric dialogue. Our core design is a progressive training approach that unifies the masked video modeling, crossmodal contrastive learning, and next token prediction, scaling up the video encoder size to 6B parameters. At the data level, we prioritize spatiotemporal consistency by semantically segmenting videos and generating video-audio-speech captions. This improves the alignment between video and text. Through extensive experiments, we validate our designs and demonstrate superior performance on over 60 video and audio tasks. Notably, our model outperforms others on various video-related dialogue and long video understanding benchmarks, highlighting its ability to reason and comprehend longer contexts. Code and models are available at https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2/.
URLs: https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2/.
Authors: Ruining Li, Chuanxia Zheng, Christian Rupprecht, Andrea Vedaldi
Abstract: We introduce DragAPart, a method that, given an image and a set of drags as input, generates a new image of the same object that responds to the action of the drags. Differently from prior works that focused on repositioning objects, DragAPart predicts part-level interactions, such as opening and closing a drawer. We study this problem as a proxy for learning a generalist motion model, not restricted to a specific kinematic structure or object category. We start from a pre-trained image generator and fine-tune it on a new synthetic dataset, Drag-a-Move, which we introduce. Combined with a new encoding for the drags and dataset randomization, the model generalizes well to real images and different categories. Compared to prior motion-controlled generators, we demonstrate much better part-level motion understanding.
Authors: Jaejung Seol, Seojun Kim, Jaejun Yoo
Abstract: Visual layout plays a critical role in graphic design fields such as advertising, posters, and web UI design. The recent trend towards content-aware layout generation through generative models has shown promise, yet it often overlooks the semantic intricacies of layout design by treating it as a simple numerical optimization. To bridge this gap, we introduce PosterLlama, a network designed for generating visually and textually coherent layouts by reformatting layout elements into HTML code and leveraging the rich design knowledge embedded within language models. Furthermore, we enhance the robustness of our model with a unique depth-based poster augmentation strategy. This ensures our generated layouts remain semantically rich but also visually appealing, even with limited data. Our extensive evaluations across several benchmarks demonstrate that PosterLlama outperforms existing methods in producing authentic and content-aware layouts. It supports an unparalleled range of conditions, including but not limited to unconditional layout generation, element conditional layout generation, layout completion, among others, serving as a highly versatile user manipulation tool.
Authors: Safouane El Ghazouali, Arnaud Gucciardi, Francesca Venturini, Nicola Venturi, Michael Rueegsegger, Umberto Michelucci
Abstract: Object detection in remotely sensed satellite pictures is fundamental in many fields such as biophysical, and environmental monitoring. While deep learning algorithms are constantly evolving, they have been mostly implemented and tested on popular ground-based taken photos. This paper critically evaluates and compares a suite of advanced object detection algorithms customized for the task of identifying aircraft within satellite imagery. Using the large HRPlanesV2 dataset, together with a rigorous validation with the GDIT dataset, this research encompasses an array of methodologies including YOLO versions 5 and 8, Faster RCNN, CenterNet, RetinaNet, RTMDet, and DETR, all trained from scratch. This exhaustive training and validation study reveal YOLOv5 as the preeminent model for the specific case of identifying airplanes from remote sensing data, showcasing high precision and adaptability across diverse imaging conditions. This research highlight the nuanced performance landscapes of these algorithms, with YOLOv5 emerging as a robust solution for aerial object detection, underlining its importance through superior mean average precision, Recall, and Intersection over Union scores. The findings described here underscore the fundamental role of algorithm selection aligned with the specific demands of satellite imagery analysis and extend a comprehensive framework to evaluate model efficacy. The benchmark toolkit and codes, available via https://github.com/toelt-llc/FlightScope_Bench, aims to further exploration and innovation in the realm of remote sensing object detection, paving the way for improved analytical methodologies in satellite imagery applications.
Authors: Mariella Dreissig, Florian Piewak, Joschka Boedecker
Abstract: Safety-critical applications such as autonomous driving require robust 3D environment perception algorithms capable of handling diverse and ambiguous surroundings. The predictive performance of classification models is heavily influenced by the dataset and the prior knowledge provided by the annotated labels. While labels guide the learning process, they often fail to capture the inherent relationships between classes that are naturally understood by humans. We propose a training strategy for a 3D LiDAR semantic segmentation model that learns structural relationships between classes through abstraction. This is achieved by implicitly modeling these relationships using a learning rule for hierarchical multi-label classification (HMC). Our detailed analysis demonstrates that this training strategy not only improves the model's confidence calibration but also retains additional information useful for downstream tasks such as fusion, prediction, and planning.
Authors: Zeyu Zhang, Xuyin Qi, Mingxi Chen, Guangxi Li, Ryan Pham, Ayub Qassim, Ella Berry, Zhibin Liao, Owen Siggs, Robert Mclaughlin, Jamie Craig, Minh-Son To
Abstract: The oxygen saturation level in the blood (SaO2) is crucial for health, particularly in relation to sleep-related breathing disorders. However, continuous monitoring of SaO2 is time-consuming and highly variable depending on patients' conditions. Recently, optical coherence tomography angiography (OCTA) has shown promising development in rapidly and effectively screening eye-related lesions, offering the potential for diagnosing sleep-related disorders. To bridge this gap, our paper presents three key contributions. Firstly, we propose JointViT, a novel model based on the Vision Transformer architecture, incorporating a joint loss function for supervision. Secondly, we introduce a balancing augmentation technique during data preprocessing to improve the model's performance, particularly on the long-tail distribution within the OCTA dataset. Lastly, through comprehensive experiments on the OCTA dataset, our proposed method significantly outperforms other state-of-the-art methods, achieving improvements of up to 12.28% in overall accuracy. This advancement lays the groundwork for the future utilization of OCTA in diagnosing sleep-related disorders. See project website https://steve-zeyu-zhang.github.io/JointViT
Authors: Vishal Nedungadi, Ankit Kariryaa, Stefan Oehmcke, Serge Belongie, Christian Igel, Nico Lang
Abstract: The volume of unlabelled Earth observation (EO) data is huge, but many important applications lack labelled training data. However, EO data offers the unique opportunity to pair data from different modalities and sensors automatically based on geographic location and time, at virtually no human labor cost. We seize this opportunity to create MMEarth, a diverse multi-modal pretraining dataset at global scale. Using this new corpus of 1.2 million locations, we propose a Multi-Pretext Masked Autoencoder (MP-MAE) approach to learn general-purpose representations for optical satellite images. Our approach builds on the ConvNeXt V2 architecture, a fully convolutional masked autoencoder (MAE). Drawing upon a suite of multi-modal pretext tasks, we demonstrate that our MP-MAE approach outperforms both MAEs pretrained on ImageNet and MAEs pretrained on domain-specific satellite images. This is shown on several downstream tasks including image classification and semantic segmentation. We find that pretraining with multi-modal pretext tasks notably improves the linear probing performance compared to pretraining on optical satellite images only. This also leads to better label efficiency and parameter efficiency which are crucial aspects in global scale applications.
Authors: Yuxiang Wei, Zhilong Ji, Jinfeng Bai, Hongzhi Zhang, Lei Zhang, Wangmeng Zuo
Abstract: Text-to-image (T2I) diffusion models have shown significant success in personalized text-to-image generation, which aims to generate novel images with human identities indicated by the reference images. Despite promising identity fidelity has been achieved by several tuning-free methods, they usually suffer from overfitting issues. The learned identity tends to entangle with irrelevant information, resulting in unsatisfied text controllability, especially on faces. In this work, we present MasterWeaver, a test-time tuning-free method designed to generate personalized images with both faithful identity fidelity and flexible editability. Specifically, MasterWeaver adopts an encoder to extract identity features and steers the image generation through additional introduced cross attention. To improve editability while maintaining identity fidelity, we propose an editing direction loss for training, which aligns the editing directions of our MasterWeaver with those of the original T2I model. Additionally, a face-augmented dataset is constructed to facilitate disentangled identity learning, and further improve the editability. Extensive experiments demonstrate that our MasterWeaver can not only generate personalized images with faithful identity, but also exhibit superiority in text controllability. Our code can be found at https://github.com/csyxwei/MasterWeaver.
Authors: Zonglin Lyu, Ming Li, Jianbo Jiao, Chen Chen
Abstract: Recent work in Video Frame Interpolation (VFI) tries to formulate VFI as a diffusion-based conditional image generation problem, synthesizing the intermediate frame given a random noise and neighboring frames. Due to the relatively high resolution of videos, Latent Diffusion Models (LDMs) are employed as the conditional generation model, where the autoencoder compresses images into latent representations for diffusion and then reconstructs images from these latent representations. Such a formulation poses a crucial challenge: VFI expects that the output is deterministically equal to the ground truth intermediate frame, but LDMs randomly generate a diverse set of different images when the model runs multiple times. The reason for the diverse generation is that the cumulative variance (variance accumulated at each step of generation) of generated latent representations in LDMs is large. This makes the sampling trajectory random, resulting in diverse rather than deterministic generations. To address this problem, we propose our unique solution: Frame Interpolation with Consecutive Brownian Bridge Diffusion. Specifically, we propose consecutive Brownian Bridge diffusion that takes a deterministic initial value as input, resulting in a much smaller cumulative variance of generated latent representations. Our experiments suggest that our method can improve together with the improvement of the autoencoder and achieve state-of-the-art performance in VFI, leaving strong potential for further enhancement.
Authors: Yihong Sun, Bharath Hariharan
Abstract: Embodied agents must detect and localize objects of interest, e.g. traffic participants for self-driving cars. Supervision in the form of bounding boxes for this task is extremely expensive. As such, prior work has looked at unsupervised instance detection and segmentation, but in the absence of annotated boxes, it is unclear how pixels must be grouped into objects and which objects are of interest. This results in over-/under-segmentation and irrelevant objects. Inspired by human visual system and practical applications, we posit that the key missing cue for unsupervised detection is motion: objects of interest are typically mobile objects that frequently move and their motions can specify separate instances. In this paper, we propose MOD-UV, a Mobile Object Detector learned from Unlabeled Videos only. We begin with instance pseudo-labels derived from motion segmentation, but introduce a novel training paradigm to progressively discover small objects and static-but-mobile objects that are missed by motion segmentation. As a result, though only learned from unlabeled videos, MOD-UV can detect and segment mobile objects from a single static image. Empirically, we achieve state-of-the-art performance in unsupervised mobile object detection on Waymo Open, nuScenes, and KITTI Datasets without using any external data or supervised models. Code is available at https://github.com/YihongSun/MOD-UV.
Authors: Bill Psomas, Ioannis Kakogeorgiou, Nikos Efthymiadis, Giorgos Tolias, Ondrej Chum, Yannis Avrithis, Konstantinos Karantzalos
Abstract: This work introduces composed image retrieval to remote sensing. It allows to query a large image archive by image examples alternated by a textual description, enriching the descriptive power over unimodal queries, either visual or textual. Various attributes can be modified by the textual part, such as shape, color, or context. A novel method fusing image-to-image and text-to-image similarity is introduced. We demonstrate that a vision-language model possesses sufficient descriptive power and no further learning step or training data are necessary. We present a new evaluation benchmark focused on color, context, density, existence, quantity, and shape modifications. Our work not only sets the state-of-the-art for this task, but also serves as a foundational step in addressing a gap in the field of remote sensing image retrieval. Code at: https://github.com/billpsomas/rscir
Authors: Mu Cai, Jianwei Yang, Jianfeng Gao, Yong Jae Lee
Abstract: Large Multimodal Models (LMMs) such as LLaVA have shown strong performance in visual-linguistic reasoning. These models first embed images into a fixed large number of visual tokens and then feed them into a Large Language Model (LLM). However, this design causes an excessive number of tokens for dense visual scenarios such as high-resolution images and videos, leading to great inefficiency. While token pruning/merging methods do exist, they produce a single length output for each image and do not afford flexibility in trading off information density v.s. efficiency. Inspired by the concept of Matryoshka Dolls, we propose M3: Matryoshka Multimodal Models, which learns to represent visual content as nested sets of visual tokens that capture information across multiple coarse-to-fine granularities. Our approach offers several unique benefits for LMMs: (1) One can explicitly control the visual granularity per test instance during inference, e.g. , adjusting the number of tokens used to represent an image based on the anticipated complexity or simplicity of the content; (2) M3 provides a framework for analyzing the granularity needed for existing datasets, where we find that COCO-style benchmarks only need around ~9 visual tokens to obtain accuracy similar to that of using all 576 tokens; (3) Our approach provides a foundation to explore the best trade-off between performance and visual token length at sample level, where our investigation reveals that a large gap exists between the oracle upper bound and current fixed-scale representations.
Authors: Yansong Qu, Shaohui Dai, Xinyang Li, Jianghang Lin, Liujuan Cao, Shengchuan Zhang, Rongrong Ji
Abstract: 3D open-vocabulary scene understanding, crucial for advancing augmented reality and robotic applications, involves interpreting and locating specific regions within a 3D space as directed by natural language instructions. To this end, we introduce GOI, a framework that integrates semantic features from 2D vision-language foundation models into 3D Gaussian Splatting (3DGS) and identifies 3D Gaussians of Interest using an Optimizable Semantic-space Hyperplane. Our approach includes an efficient compression method that utilizes scene priors to condense noisy high-dimensional semantic features into compact low-dimensional vectors, which are subsequently embedded in 3DGS. During the open-vocabulary querying process, we adopt a distinct approach compared to existing methods, which depend on a manually set fixed empirical threshold to select regions based on their semantic feature distance to the query text embedding. This traditional approach often lacks universal accuracy, leading to challenges in precisely identifying specific target areas. Instead, our method treats the feature selection process as a hyperplane division within the feature space, retaining only those features that are highly relevant to the query. We leverage off-the-shelf 2D Referring Expression Segmentation (RES) models to fine-tune the semantic-space hyperplane, enabling a more precise distinction between target regions and others. This fine-tuning substantially improves the accuracy of open-vocabulary queries, ensuring the precise localization of pertinent 3D Gaussians. Extensive experiments demonstrate GOI's superiority over previous state-of-the-art methods. Our project page is available at https://quyans.github.io/GOI-Hyperplane/ .
Authors: Vitor Fortes Rey, Lala Shakti Swarup Ray, Xia Qingxin, Kaishun Wu, Paul Lukowicz
Abstract: Due to the scarcity of labeled sensor data in HAR, prior research has turned to video data to synthesize Inertial Measurement Units (IMU) data, capitalizing on its rich activity annotations. However, generating IMU data from videos presents challenges for HAR in real-world settings, attributed to the poor quality of synthetic IMU data and its limited efficacy in subtle, fine-grained motions. In this paper, we propose Multi$^3$Net, our novel multi-modal, multitask, and contrastive-based framework approach to address the issue of limited data. Our pretraining procedure uses videos from online repositories, aiming to learn joint representations of text, pose, and IMU simultaneously. By employing video data and contrastive learning, our method seeks to enhance wearable HAR performance, especially in recognizing subtle activities.Our experimental findings validate the effectiveness of our approach in improving HAR performance with IMU data. We demonstrate that models trained with synthetic IMU data generated from videos using our method surpass existing approaches in recognizing fine-grained activities.
Authors: Giuseppe Vecchio
Abstract: We introduce StableMaterials, a novel approach for generating photorealistic physical-based rendering (PBR) materials that integrate semi-supervised learning with Latent Diffusion Models (LDMs). Our method employs adversarial training to distill knowledge from existing large-scale image generation models, minimizing the reliance on annotated data and enhancing the diversity in generation. This distillation approach aligns the distribution of the generated materials with that of image textures from an SDXL model, enabling the generation of novel materials that are not present in the initial training dataset. Furthermore, we employ a diffusion-based refiner model to improve the visual quality of the samples and achieve high-resolution generation. Finally, we distill a latent consistency model for fast generation in just four steps and propose a new tileability technique that removes visual artifacts typically associated with fewer diffusion steps. We detail the architecture and training process of StableMaterials, the integration of semi-supervised training within existing LDM frameworks and show the advantages of our approach. Comparative evaluations with state-of-the-art methods show the effectiveness of StableMaterials, highlighting its potential applications in computer graphics and beyond. StableMaterials is publicly available at https://gvecchio.com/stablematerials.
Authors: Tianhong Li, Yonglong Tian, He Li, Mingyang Deng, Kaiming He
Abstract: Conventional wisdom holds that autoregressive models for image generation are typically accompanied by vector-quantized tokens. We observe that while a discrete-valued space can facilitate representing a categorical distribution, it is not a necessity for autoregressive modeling. In this work, we propose to model the per-token probability distribution using a diffusion procedure, which allows us to apply autoregressive models in a continuous-valued space. Rather than using categorical cross-entropy loss, we define a Diffusion Loss function to model the per-token probability. This approach eliminates the need for discrete-valued tokenizers. We evaluate its effectiveness across a wide range of cases, including standard autoregressive models and generalized masked autoregressive (MAR) variants. By removing vector quantization, our image generator achieves strong results while enjoying the speed advantage of sequence modeling. We hope this work will motivate the use of autoregressive generation in other continuous-valued domains and applications. Code is available at: https://github.com/LTH14/mar
Authors: Zhuoxiao Chen, Zixin Wang, Yadan Luo, Sen Wang, Zi Huang
Abstract: LiDAR-based 3D object detection has seen impressive advances in recent times. However, deploying trained 3D detectors in the real world often yields unsatisfactory performance when the distribution of the test data significantly deviates from the training data due to different weather conditions, object sizes, \textit{etc}. A key factor in this performance degradation is the diminished generalizability of pre-trained models, which creates a sharp loss landscape during training. Such sharpness, when encountered during testing, can precipitate significant performance declines, even with minor data variations. To address the aforementioned challenges, we propose \textbf{dual-perturbation optimization (DPO)} for \textbf{\underline{T}est-\underline{t}ime \underline{A}daptation in \underline{3}D \underline{O}bject \underline{D}etection (TTA-3OD)}. We minimize the sharpness to cultivate a flat loss landscape to ensure model resiliency to minor data variations, thereby enhancing the generalization of the adaptation process. To fully capture the inherent variability of the test point clouds, we further introduce adversarial perturbation to the input BEV features to better simulate the noisy test environment. As the dual perturbation strategy relies on trustworthy supervision signals, we utilize a reliable Hungarian matcher to filter out pseudo-labels sensitive to perturbations. Additionally, we introduce early Hungarian cutoff to avoid error accumulation from incorrect pseudo-labels by halting the adaptation process. Extensive experiments across three types of transfer tasks demonstrate that the proposed DPO significantly surpasses previous state-of-the-art approaches, specifically on Waymo $\rightarrow$ KITTI, outperforming the most competitive baseline by 57.72\% in $\text{AP}_\text{3D}$ and reaching 91\% of the fully supervised upper bound.
Authors: Guanghao Zhu, Lin Liu, Yuhao Hu, Haixin Sun, Fang Liu, Xiaohui Du, Ruqian Hao, Juanxiu Liu, Yong Liu, Hao Deng, Jing Zhang
Abstract: Micro-expressions are subtle facial movements that occur spontaneously when people try to conceal real emotions. Micro-expression recognition is crucial in many fields, including criminal analysis and psychotherapy. However, micro-expression recognition is challenging since micro-expressions have low intensity and public datasets are small in size. To this end, a three-stream temporal-shift attention network based on self-knowledge distillation called SKD-TSTSAN is proposed in this paper. Firstly, to address the low intensity of muscle movements, we utilize learning-based motion magnification modules to enhance the intensity of muscle movements. Secondly, we employ efficient channel attention modules in the local-spatial stream to make the network focus on facial regions that are highly relevant to micro-expressions. In addition, temporal shift modules are used in the dynamic-temporal stream, which enables temporal modeling with no additional parameters by mixing motion information from two different temporal domains. Furthermore, we introduce self-knowledge distillation into the micro-expression recognition task by introducing auxiliary classifiers and using the deepest section of the network for supervision, encouraging all blocks to fully explore the features of the training set. Finally, extensive experiments are conducted on four public datasets: CASME II, SAMM, MMEW, and CAS(ME)3. The experimental results demonstrate that our SKD-TSTSAN outperforms other existing methods and achieves new state-of-the-art performance. Our code will be available at https://github.com/GuanghaoZhu663/SKD-TSTSAN.
Authors: Muhammad Salman Ali, Maryam Qamar, Sung-Ho Bae, Enzo Tartaglione
Abstract: In recent times, the utilization of 3D models has gained traction, owing to the capacity for end-to-end training initially offered by Neural Radiance Fields and more recently by 3D Gaussian Splatting (3DGS) models. The latter holds a significant advantage by inherently easing rapid convergence during training and offering extensive editability. However, despite rapid advancements, the literature still lives in its infancy regarding the scalability of these models. In this study, we take some initial steps in addressing this gap, showing an approach that enables both the memory and computational scalability of such models. Specifically, we propose "Trimming the fat", a post-hoc gradient-informed iterative pruning technique to eliminate redundant information encoded in the model. Our experimental findings on widely acknowledged benchmarks attest to the effectiveness of our approach, revealing that up to 75% of the Gaussians can be removed while maintaining or even improving upon baseline performance. Our approach achieves around 50$\times$ compression while preserving performance similar to the baseline model, and is able to speed-up computation up to 600 FPS.
Authors: Weihong Zhong, Xiaocheng Feng, Liang Zhao, Qiming Li, Lei Huang, Yuxuan Gu, Weitao Ma, Yuan Xu, Bing Qin
Abstract: Though advanced in understanding visual information with human languages, Large Vision-Language Models (LVLMs) still suffer from multimodal hallucinations. A natural concern is that during multimodal interaction, the generated hallucinations could influence the LVLMs' subsequent generation. Thus, we raise a question: When presented with a query relevant to the previously generated hallucination, will LVLMs be misled and respond incorrectly, even though the ground visual information exists? To answer this, we propose a framework called MMHalSnowball to evaluate LVLMs' behaviors when encountering generated hallucinations, where LVLMs are required to answer specific visual questions within a curated hallucinatory conversation. Crucially, our experiment shows that the performance of open-source LVLMs drops by at least $31\%$, indicating that LVLMs are prone to accept the generated hallucinations and make false claims that they would not have supported without distractions. We term this phenomenon Multimodal Hallucination Snowballing. To mitigate this, we further propose a training-free method called Residual Visual Decoding, where we revise the output distribution of LVLMs with the one derived from the residual visual input, providing models with direct access to the visual information. Experiments show that our method can mitigate more than $24\%$ of the snowballed multimodal hallucination while maintaining capabilities.
Authors: Ruirui Lin, Nantheera Anantrasirichai, Guoxi Huang, Joanne Lin, Qi Sun, Alexandra Malyugina, David R Bull
Abstract: Low-light videos often exhibit spatiotemporal incoherent noise, compromising visibility and performance in computer vision applications. One significant challenge in enhancing such content using deep learning is the scarcity of training data. This paper introduces a novel low-light video dataset, consisting of 40 scenes with various motion scenarios under two distinct low-lighting conditions, incorporating genuine noise and temporal artifacts. We provide fully registered ground truth data captured in normal light using a programmable motorized dolly and refine it via an image-based approach for pixel-wise frame alignment across different light levels. We provide benchmarks based on four different technologies: convolutional neural networks, transformers, diffusion models, and state space models (mamba). Our experimental results demonstrate the significance of fully registered video pairs for low-light video enhancement (LLVE) and the comprehensive evaluation shows that the models trained with our dataset outperform those trained with the existing datasets. Our dataset and links to benchmarks are publicly available at https://doi.org/10.21227/mzny-8c77.
Authors: Kai Norman Clasen, Leonard Hackel, Tom Burgert, Gencer Sumbul, Beg\"um Demir, Volker Markl
Abstract: This paper presents refined BigEarthNet (reBEN) that is a large-scale, multi-modal remote sensing dataset constructed to support deep learning (DL) studies for remote sensing image analysis. The reBEN dataset consists of 549,488 pairs of Sentinel-1 and Sentinel-2 image patches. To construct reBEN, we initially consider the Sentinel-1 and Sentinel-2 tiles used to construct the BigEarthNet dataset and then divide them into patches of size 1200 m x 1200 m. We apply atmospheric correction to the Sentinel-2 patches using the latest version of the sen2cor tool, resulting in higher-quality patches compared to those present in BigEarthNet. Each patch is then associated with a pixel-level reference map and scene-level multi-labels. This makes reBEN suitable for pixel- and scene-based learning tasks. The labels are derived from the most recent CORINE Land Cover (CLC) map of 2018 by utilizing the 19-class nomenclature as in BigEarthNet. The use of the most recent CLC map results in overcoming the label noise present in BigEarthNet. Furthermore, we introduce a new geographical-based split assignment algorithm that significantly reduces the spatial correlation among the train, validation, and test sets with respect to those present in BigEarthNet. This increases the reliability of the evaluation of DL models. To minimize the DL model training time, we introduce software tools that convert the reBEN dataset into a DL-optimized data format. In our experiments, we show the potential of reBEN for multi-modal multi-label image classification problems by considering several state-of-the-art DL models. The pre-trained model weights, associated code, and complete dataset are available at https://bigearth.net.
URLs: https://bigearth.net.
Authors: Feng Li, Renrui Zhang, Hao Zhang, Yuanhan Zhang, Bo Li, Wei Li, Zejun Ma, Chunyuan Li
Abstract: Visual instruction tuning has made considerable strides in enhancing the capabilities of Large Multimodal Models (LMMs). However, existing open LMMs largely focus on single-image tasks, their applications to multi-image scenarios remains less explored. Additionally, prior LMM research separately tackles different scenarios, leaving it impossible to generalize cross scenarios with new emerging capabilities. To this end, we introduce LLaVA-NeXT-Interleave, which simultaneously tackles Multi-image, Multi-frame (video), Multi-view (3D), and Multi-patch (single-image) scenarios in LMMs. To enable these capabilities, we regard the interleaved data format as a general template and compile the M4-Instruct dataset with 1,177.6k samples, spanning 4 primary domains with 14 tasks and 41 datasets. We also curate the LLaVA-Interleave Bench to comprehensively evaluate the multi-image performance of LMMs. Through extensive experiments, LLaVA-NeXT-Interleave achieves leading results in multi-image, video, and 3D benchmarks, while maintaining the performance of single-image tasks. Besides, our model also exhibits several emerging capabilities, e.g., transferring tasks across different settings and modalities. Code is available at https://github.com/LLaVA-VL/LLaVA-NeXT
Authors: Ningli Xu, Rongjun Qin
Abstract: Predicting realistic ground views from satellite imagery in urban scenes is a challenging task due to the significant view gaps between satellite and ground-view images. We propose a novel pipeline to tackle this challenge, by generating geospecifc views that maximally respect the weak geometry and texture from multi-view satellite images. Different from existing approaches that hallucinate images from cues such as partial semantics or geometry from overhead satellite images, our method directly predicts ground-view images at geolocation by using a comprehensive set of information from the satellite image, resulting in ground-level images with a resolution boost at a factor of ten or more. We leverage a novel building refinement method to reduce geometric distortions in satellite data at ground level, which ensures the creation of accurate conditions for view synthesis using diffusion networks. Moreover, we proposed a novel geospecific prior, which prompts distribution learning of diffusion models to respect image samples that are closer to the geolocation of the predicted images. We demonstrate our pipeline is the first to generate close-to-real and geospecific ground views merely based on satellite images.
Authors: Zujin Guo, Wei Li, Chen Change Loy
Abstract: Motion modeling is critical in flow-based Video Frame Interpolation (VFI). Existing paradigms either consider linear combinations of bidirectional flows or directly predict bilateral flows for given timestamps without exploring favorable motion priors, thus lacking the capability of effectively modeling spatiotemporal dynamics in real-world videos. To address this limitation, in this study, we introduce Generalizable Implicit Motion Modeling (GIMM), a novel and effective approach to motion modeling for VFI. Specifically, to enable GIMM as an effective motion modeling paradigm, we design a motion encoding pipeline to model spatiotemporal motion latent from bidirectional flows extracted from pre-trained flow estimators, effectively representing input-specific motion priors. Then, we implicitly predict arbitrary-timestep optical flows within two adjacent input frames via an adaptive coordinate-based neural network, with spatiotemporal coordinates and motion latent as inputs. Our GIMM can be smoothly integrated with existing flow-based VFI works without further modifications. We show that GIMM performs better than the current state of the art on the VFI benchmarks.
Authors: Chen Xin, Andreas Hartel, Enkelejda Kasneci
Abstract: Accurate real-time object detection is vital across numerous industrial applications, from safety monitoring to quality control. Traditional approaches, however, are hindered by arduous manual annotation and data collection, struggling to adapt to ever-changing environments and novel target objects. To address these limitations, this paper presents DART, an innovative automated end-to-end pipeline that revolutionizes object detection workflows from data collection to model evaluation. It eliminates the need for laborious human labeling and extensive data collection while achieving outstanding accuracy across diverse scenarios. DART encompasses four key stages: (1) Data Diversification using subject-driven image generation (DreamBooth with SDXL), (2) Annotation via open-vocabulary object detection (Grounding DINO) to generate bounding box and class labels, (3) Review of generated images and pseudo-labels by large multimodal models (InternVL-1.5 and GPT-4o) to guarantee credibility, and (4) Training of real-time object detectors (YOLOv8 and YOLOv10) using the verified data. We apply DART to a self-collected dataset of construction machines named Liebherr Product, which contains over 15K high-quality images across 23 categories. The current instantiation of DART significantly increases average precision (AP) from 0.064 to 0.832. Its modular design ensures easy exchangeability and extensibility, allowing for future algorithm upgrades, seamless integration of new object categories, and adaptability to customized environments without manual labeling and additional data collection. The code and dataset are released at https://github.com/chen-xin-94/DART.
Authors: Hongsong Wang, Jianhua Zhao, Jie Gui
Abstract: Human action understanding is a fundamental and challenging task in computer vision. Although there exists tremendous research on this area, most works focus on action recognition, while action retrieval has received less attention. In this paper, we focus on the neglected but important task of image-based action retrieval which aims to find images that depict the same action as a query image. We establish benchmarks for this task and set up important baseline methods for fair comparison. We present an end-to-end model that learns rich action representations from three aspects: the anchored person, contextual regions, and the global image. A novel fusion transformer module is designed to model the relationships among different features and effectively fuse them into an action representation. Experiments on the Stanford-40 and PASCAL VOC 2012 Action datasets show that the proposed method significantly outperforms previous approaches for image-based action retrieval.
Authors: Yixiang Qiu, Hao Fang, Hongyao Yu, Bin Chen, MeiKang Qiu, Shu-Tao Xia
Abstract: Model Inversion (MI) attacks aim to reconstruct privacy-sensitive training data from released models by utilizing output information, raising extensive concerns about the security of Deep Neural Networks (DNNs). Recent advances in generative adversarial networks (GANs) have contributed significantly to the improved performance of MI attacks due to their powerful ability to generate realistic images with high fidelity and appropriate semantics. However, previous MI attacks have solely disclosed private information in the latent space of GAN priors, limiting their semantic extraction and transferability across multiple target models and datasets. To address this challenge, we propose a novel method, Intermediate Features enhanced Generative Model Inversion (IF-GMI), which disassembles the GAN structure and exploits features between intermediate blocks. This allows us to extend the optimization space from latent code to intermediate features with enhanced expressive capabilities. To prevent GAN priors from generating unrealistic images, we apply a L1 ball constraint to the optimization process. Experiments on multiple benchmarks demonstrate that our method significantly outperforms previous approaches and achieves state-of-the-art results under various settings, especially in the out-of-distribution (OOD) scenario. Our code is available at: https://github.com/final-solution/IF-GMI
Authors: Zipeng Qi, Lichen Bai, Haoyi Xiong, Zeke Xie
Abstract: Diffusion models that can generate high-quality data from randomly sampled Gaussian noises have become the mainstream generative method in both academia and industry. Are randomly sampled Gaussian noises equally good for diffusion models? While a large body of works tried to understand and improve diffusion models, previous works overlooked the possibility to select or optimize the sampled noise the possibility of selecting or optimizing sampled noises for improving diffusion models. In this paper, we mainly made three contributions. First, we report that not all noises are created equally for diffusion models. We are the first to hypothesize and empirically observe that the generation quality of diffusion models significantly depend on the noise inversion stability. This naturally provides us a noise selection method according to the inversion stability. Second, we further propose a novel noise optimization method that actively enhances the inversion stability of arbitrary given noises. Our method is the first one that works on noise space to generally improve generated results without fine-tuning diffusion models. Third, our extensive experiments demonstrate that the proposed noise selection and noise optimization methods both significantly improve representative diffusion models, such as SDXL and SDXL-turbo, in terms of human preference and other objective evaluation metrics. For example, the human preference winning rates of noise selection and noise optimization over the baselines can be up to 57% and 72.5%, respectively, on DrawBench.
Authors: Rongkun Zheng, Lu Qi, Xi Chen, Yi Wang, Kun Wang, Yu Qiao, Hengshuang Zhao
Abstract: Although video perception models have made remarkable advancements in recent years, they still heavily rely on explicit text descriptions or pre-defined categories to identify target instances before executing video perception tasks. These models, however, fail to proactively comprehend and reason the user's intentions via textual input. Even though previous works attempt to investigate solutions to incorporate reasoning with image segmentation, they fail to reason with videos due to the video's complexity in object motion. To bridge the gap between image and video, in this work, we propose a new video segmentation task - video reasoning segmentation. The task is designed to output tracklets of segmentation masks given a complex input text query. What's more, to promote research in this unexplored area, we construct a reasoning video segmentation benchmark. Finally, we present ViLLa: Video reasoning segmentation with a Large Language Model, which incorporates the language generation capabilities of multimodal Large Language Models (LLMs) while retaining the capabilities of detecting, segmenting, and tracking multiple instances. We use a temporal-aware context aggregation module to incorporate contextual visual cues to text embeddings and propose a video-frame decoder to build temporal correlations across segmentation tokens. Remarkably, our ViLLa demonstrates capability in handling complex reasoning and referring video segmentation. Also, our model shows impressive ability in different temporal understanding benchmarks. Both quantitative and qualitative experiments show our method effectively unlocks new video reasoning segmentation capabilities for multimodal LLMs. The code and dataset will be available at https://github.com/rkzheng99/ViLLa.
Authors: Cuong Pham, Hoang Anh Dung, Cuong C. Nguyen, Trung Le, Dinh Phung, Gustavo Carneiro, Thanh-Toan Do
Abstract: Post-Training Quantization (PTQ) has received significant attention because it requires only a small set of calibration data to quantize a full-precision model, which is more practical in real-world applications in which full access to a large training set is not available. However, it often leads to overfitting on the small calibration dataset. Several methods have been proposed to address this issue, yet they still rely on only the calibration set for the quantization and they do not validate the quantized model due to the lack of a validation set. In this work, we propose a novel meta-learning based approach to enhance the performance of post-training quantization. Specifically, to mitigate the overfitting problem, instead of only training the quantized model using the original calibration set without any validation during the learning process as in previous PTQ works, in our approach, we both train and validate the quantized model using two different sets of images. In particular, we propose a meta-learning based approach to jointly optimize a transformation network and a quantized model through bi-level optimization. The transformation network modifies the original calibration data and the modified data will be used as the training set to learn the quantized model with the objective that the quantized model achieves a good performance on the original calibration data. Extensive experiments on the widely used ImageNet dataset with different neural network architectures demonstrate that our approach outperforms the state-of-the-art PTQ methods.
Authors: Xiaomeng Chu, Jiajun Deng, Guoliang You, Yifan Duan, Yao Li, Yanyong Zhang
Abstract: The recent advances in query-based multi-camera 3D object detection are featured by initializing object queries in the 3D space, and then sampling features from perspective-view images to perform multi-round query refinement. In such a framework, query points near the same camera ray are likely to sample similar features from very close pixels, resulting in ambiguous query features and degraded detection accuracy. To this end, we introduce RayFormer, a camera-ray-inspired query-based 3D object detector that aligns the initialization and feature extraction of object queries with the optical characteristics of cameras. Specifically, RayFormer transforms perspective-view image features into bird's eye view (BEV) via the lift-splat-shoot method and segments the BEV map to sectors based on the camera rays. Object queries are uniformly and sparsely initialized along each camera ray, facilitating the projection of different queries onto different areas in the image to extract distinct features. Besides, we leverage the instance information of images to supplement the uniformly initialized object queries by further involving additional queries along the ray from 2D object detection boxes. To extract unique object-level features that cater to distinct queries, we design a ray sampling method that suitably organizes the distribution of feature sampling points on both images and bird's eye view. Extensive experiments are conducted on the nuScenes dataset to validate our proposed ray-inspired model design. The proposed RayFormer achieves 55.5% mAP and 63.3% NDS, respectively. Our codes will be made available.
Authors: Zeyu Wang, Jingyu Lin, Yifei Qian, Yi Huang, Shicen Tian, Bosong Chai, Juncan Deng, Lan Du, Cunjian Chen, Yufei Guo, Kejie Huang
Abstract: Diffusion models have made significant strides in language-driven and layout-driven image generation. However, most diffusion models are limited to visible RGB image generation. In fact, human perception of the world is enriched by diverse viewpoints, including chromatic contrast, thermal illumination, and depth information. In this paper, we introduce a novel diffusion model for general layout-guided cross-modal ``RGB+X'' generation, called DiffX. Firstly, we construct the cross-modal image datasets with text description by using LLaVA for image captioning, supplemented by manual corrections. Notably, DiffX presents a simple yet effective cross-modal generative modeling pipeline, which conducts diffusion and denoising processes in the modality-shared latent space, facilitated by our Dual Path Variational AutoEncoder (DP-VAE). Moreover, we introduce the joint-modality embedder, which incorporates a gated cross-attention mechanism to link layout and text conditions. Meanwhile, the advanced Long-CLIP is employed for long caption embedding to improve user guidance. Through extensive experiments, DiffX demonstrates robustness and flexibility in cross-modal generation across three RGB+X datasets: FLIR, MFNet, and COME15K, guided by various layout types. It also shows the potential for adaptive generation of ``RGB+X+Y'' or more diverse modalities. Our code and constructed cross-modal image datasets are available at https://github.com/zeyuwang-zju/DiffX.
Authors: Xi Shi, Lingli Chen, Peng Wei, Xi Wu, Tian Jiang, Yonggang Luo, Lecheng Xie
Abstract: Existing Gaussian splatting methods often fall short in achieving satisfactory novel view synthesis in driving scenes, primarily due to the absence of crafty design and geometric constraints for the involved elements. This paper introduces a novel neural rendering method termed Decoupled Hybrid Gaussian Splatting (DHGS), targeting at promoting the rendering quality of novel view synthesis for static driving scenes. The novelty of this work lies in the decoupled and hybrid pixel-level blender for road and non-road layers, without the conventional unified differentiable rendering logic for the entire scene, while still maintaining consistent and continuous superimposition through the proposed depth-ordered hybrid rendering strategy. Additionally, an implicit road representation comprised of a Signed Distance Field (SDF) is trained to supervise the road surface with subtle geometric attributes. Accompanied by the use of auxiliary transmittance loss and consistency loss, novel images with imperceptible boundary and elevated fidelity are ultimately obtained. Substantial experiments on the Waymo dataset prove that DHGS outperforms the state-of-the-art methods. The project page where more video evidences are given is: https://ironbrotherstyle.github.io/dhgs_web.
Authors: Zhenzhi Wang, Yixuan Li, Yanhong Zeng, Youqing Fang, Yuwei Guo, Wenran Liu, Jing Tan, Kai Chen, Tianfan Xue, Bo Dai, Dahua Lin
Abstract: Human image animation involves generating videos from a character photo, allowing user control and unlocking potential for video and movie production. While recent approaches yield impressive results using high-quality training data, the inaccessibility of these datasets hampers fair and transparent benchmarking. Moreover, these approaches prioritize 2D human motion and overlook the significance of camera motions in videos, leading to limited control and unstable video generation. To demystify the training data, we present HumanVid, the first large-scale high-quality dataset tailored for human image animation, which combines crafted real-world and synthetic data. For the real-world data, we compile a vast collection of copyright-free real-world videos from the internet. Through a carefully designed rule-based filtering strategy, we ensure the inclusion of high-quality videos, resulting in a collection of 20K human-centric videos in 1080P resolution. Human and camera motion annotation is accomplished using a 2D pose estimator and a SLAM-based method. For the synthetic data, we gather 2,300 copyright-free 3D avatar assets to augment existing available 3D assets. Notably, we introduce a rule-based camera trajectory generation method, enabling the synthetic pipeline to incorporate diverse and precise camera motion annotation, which can rarely be found in real-world data. To verify the effectiveness of HumanVid, we establish a baseline model named CamAnimate, short for Camera-controllable Human Animation, that considers both human and camera motions as conditions. Through extensive experimentation, we demonstrate that such simple baseline training on our HumanVid achieves state-of-the-art performance in controlling both human pose and camera motions, setting a new benchmark. Code and data will be publicly available at https://github.com/zhenzhiwang/HumanVid/.
Authors: Shougao Zhang, Mengqi Zhou, Yuxi Wang, Chuanchen Luo, Rongyu Wang, Yiwei Li, Xucheng Yin, Zhaoxiang Zhang, Junran Peng
Abstract: Generating a realistic, large-scale 3D virtual city remains a complex challenge due to the involvement of numerous 3D assets, various city styles, and strict layout constraints. Existing approaches provide promising attempts at procedural content generation to create large-scale scenes using Blender agents. However, they face crucial issues such as difficulties in scaling up generation capability and achieving fine-grained control at the semantic layout level. To address these problems, we propose a novel multi-modal controllable procedural content generation method, named CityX, which enhances realistic, unbounded 3D city generation guided by multiple layout conditions, including OSM, semantic maps, and satellite images. Specifically, the proposed method contains a general protocol for integrating various PCG plugins and a multi-agent framework for transforming instructions into executable Blender actions. Through this effective framework, CityX shows the potential to build an innovative ecosystem for 3D scene generation by bridging the gap between the quality of generated assets and industrial requirements. Extensive experiments have demonstrated the effectiveness of our method in creating high-quality, diverse, and unbounded cities guided by multi-modal conditions. Our project page: https://cityx-lab.github.io.
Authors: Nuria Alabau-Bosque, Paula Daud\'en-Oliver, Jorge Vila-Tom\'as, Valero Laparra, Jes\'us Malo
Abstract: Deep architectures are the current state-of-the-art in predicting subjective image quality. Usually, these models are evaluated according to their ability to correlate with human opinion in databases with a range of distortions that may appear in digital media. However, these oversee affine transformations which may represent better the changes in the images actually happening in natural conditions. Humans can be particularly invariant to these natural transformations, as opposed to the digital ones. In this work, we evaluate state-of-the-art deep image quality metrics by assessing their invariance to affine transformations, specifically: rotation, translation, scaling, and changes in spectral illumination. Here invariance of a metric refers to the fact that certain distances should be neglected (considered to be zero) if their values are below a threshold. This is what we call invisibility threshold of a metric. We propose a methodology to assign such invisibility thresholds for any perceptual metric. This methodology involves transformations to a distance space common to any metric, and psychophysical measurements of thresholds in this common space. By doing so, we allow the analyzed metrics to be directly comparable with actual human thresholds. We find that none of the state-of-the-art metrics shows human-like results under this strong test based on invisibility thresholds. This means that tuning the models exclusively to predict the visibility of generic distortions may disregard other properties of human vision as for instance invariances or invisibility thresholds.
Authors: Yu-Kai Huang, Yutong Zheng, Yen-Shuo Su, Anudeepsekhar Bolimera, Han Zhang, Fangyi Chen, Marios Savvides
Abstract: Facial attribute editing plays a crucial role in synthesizing realistic faces with specific characteristics while maintaining realistic appearances. Despite advancements, challenges persist in achieving precise, 3D-aware attribute modifications, which are crucial for consistent and accurate representations of faces from different angles. Current methods struggle with semantic entanglement and lack effective guidance for incorporating attributes while maintaining image integrity. To address these issues, we introduce a novel framework that merges the strengths of latent-based and reference-based editing methods. Our approach employs a 3D GAN inversion technique to embed attributes from the reference image into a tri-plane space, ensuring 3D consistency and realistic viewing from multiple perspectives. We utilize blending techniques and predicted semantic masks to locate precise edit regions, merging them with the contextual guidance from the reference image. A coarse-to-fine inpainting strategy is then applied to preserve the integrity of untargeted areas, significantly enhancing realism. Our evaluations demonstrate superior performance across diverse editing tasks, validating our framework's effectiveness in realistic and applicable facial attribute editing.
Authors: Jingjun Yi, Qi Bi, Hao Zheng, Haolan Zhan, Wei Ji, Yawen Huang, Yuexiang Li, Yefeng Zheng
Abstract: The rapid development of Vision Foundation Model (VFM) brings inherent out-domain generalization for a variety of down-stream tasks. Among them, domain generalized semantic segmentation (DGSS) holds unique challenges as the cross-domain images share common pixel-wise content information but vary greatly in terms of the style. In this paper, we present a novel Spectral-dEcomposed Token (SET) learning framework to advance the frontier. Delving into further than existing fine-tuning token & frozen backbone paradigm, the proposed SET especially focuses on the way learning style-invariant features from these learnable tokens. Particularly, the frozen VFM features are first decomposed into the phase and amplitude components in the frequency space, which mainly contain the information of content and style, respectively, and then separately processed by learnable tokens for task-specific information extraction. After the decomposition, style variation primarily impacts the token-based feature enhancement within the amplitude branch. To address this issue, we further develop an attention optimization method to bridge the gap between style-affected representation and static tokens during inference. Extensive cross-domain experiments show its state-of-the-art performance.
Authors: In Cho, Hyunbo Shim, Seon Joo Kim
Abstract: This paper aims to facilitate more practical NLOS imaging by reducing the number of samplings and scan areas. To this end, we introduce a phasor-based enhancement network that is capable of predicting clean and full measurements from noisy partial observations. We leverage a denoising autoencoder scheme to acquire rich and noise-robust representations in the measurement space. Through this pipeline, our enhancement network is trained to accurately reconstruct complete measurements from their corrupted and partial counterparts. However, we observe that the \naive application of denoising often yields degraded and over-smoothed results, caused by unnecessary and spurious frequency signals present in measurements. To address this issue, we introduce a phasor-based pipeline designed to limit the spectrum of our network to the frequency range of interests, where the majority of informative signals are detected. The phasor wavefronts at the aperture, which are band-limited signals, are employed as inputs and outputs of the network, guiding our network to learn from the frequency range of interests and discard unnecessary information. The experimental results in more practical acquisition scenarios demonstrate that we can look around the corners with $16\times$ or $64\times$ fewer samplings and $4\times$ smaller apertures. Our code is available at https://github.com/join16/LEAP.
Authors: Sarah M\"uller, Louisa Fay, Lisa M. Koch, Sergios Gatidis, Thomas K\"ustner, Philipp Berens
Abstract: Medical imaging cohorts are often confounded by factors such as acquisition devices, hospital sites, patient backgrounds, and many more. As a result, deep learning models tend to learn spurious correlations instead of causally related features, limiting their generalizability to new and unseen data. This problem can be addressed by minimizing dependence measures between intermediate representations of task-related and non-task-related variables. These measures include mutual information, distance correlation, and the performance of adversarial classifiers. Here, we benchmark such dependence measures for the task of preventing shortcut learning. We study a simplified setting using Morpho-MNIST and a medical imaging task with CheXpert chest radiographs. Our results provide insights into how to mitigate confounding factors in medical imaging.
Authors: Stefan Thalhammer, Dominik Bauer, Peter H\"onig, Jean-Baptiste Weibel, Jos\'e Garc\'ia-Rodr\'iguez, Markus Vincze
Abstract: Object pose estimation is a core perception task that enables, for example, object grasping and scene understanding. The widely available, inexpensive and high-resolution RGB sensors and CNNs that allow for fast inference based on this modality make monocular approaches especially well suited for robotics applications. We observe that previous surveys on object pose estimation establish the state of the art for varying modalities, single- and multi-view settings, and datasets and metrics that consider a multitude of applications. We argue, however, that those works' broad scope hinders the identification of open challenges that are specific to monocular approaches and the derivation of promising future challenges for their application in robotics. By providing a unified view on recent publications from both robotics and computer vision, we find that occlusion handling, novel pose representations, and formalizing and improving category-level pose estimation are still fundamental challenges that are highly relevant for robotics. Moreover, to further improve robotic performance, large object sets, novel objects, refractive materials, and uncertainty estimates are central, largely unsolved open challenges. In order to address them, ontological reasoning, deformability handling, scene-level reasoning, realistic datasets, and the ecological footprint of algorithms need to be improved.
Authors: Yanjie Ze, Ge Yan, Yueh-Hua Wu, Annabella Macaluso, Yuying Ge, Jianglong Ye, Nicklas Hansen, Li Erran Li, Xiaolong Wang
Abstract: It is a long-standing problem in robotics to develop agents capable of executing diverse manipulation tasks from visual observations in unstructured real-world environments. To achieve this goal, the robot needs to have a comprehensive understanding of the 3D structure and semantics of the scene. In this work, we present $\textbf{GNFactor}$, a visual behavior cloning agent for multi-task robotic manipulation with $\textbf{G}$eneralizable $\textbf{N}$eural feature $\textbf{F}$ields. GNFactor jointly optimizes a generalizable neural field (GNF) as a reconstruction module and a Perceiver Transformer as a decision-making module, leveraging a shared deep 3D voxel representation. To incorporate semantics in 3D, the reconstruction module utilizes a vision-language foundation model ($\textit{e.g.}$, Stable Diffusion) to distill rich semantic information into the deep 3D voxel. We evaluate GNFactor on 3 real robot tasks and perform detailed ablations on 10 RLBench tasks with a limited number of demonstrations. We observe a substantial improvement of GNFactor over current state-of-the-art methods in seen and unseen tasks, demonstrating the strong generalization ability of GNFactor. Our project website is https://yanjieze.com/GNFactor/ .
Authors: Imad Eddine Marouf, Subhankar Roy, Enzo Tartaglione, St\'ephane Lathuili\`ere
Abstract: In this work, we study the problem of continual learning (CL) where the goal is to learn a model on a sequence of tasks, such that the data from the previous tasks becomes unavailable while learning on the current task data. CL is essentially a balancing act between being able to learn on the new task (i.e., plasticity) and maintaining the performance on the previously learned concepts (i.e., stability). Intending to address the stability-plasticity trade-off, we propose to perform weight-ensembling of the model parameters of the previous and current tasks. This weighted-ensembled model, which we call Continual Model Averaging (or CoMA), attains high accuracy on the current task by leveraging plasticity, while not deviating too far from the previous weight configuration, ensuring stability. We also propose an improved variant of CoMA, named Continual Fisher-weighted Model Averaging (or CoFiMA), that selectively weighs each parameter in the weights ensemble by leveraging the Fisher information of the weights of the model. Both variants are conceptually simple, easy to implement, and effective in attaining state-of-the-art performance on several standard CL benchmarks. Code is available at: https://github.com/IemProg/CoFiMA.
Authors: Yunfeng Fan, Wenchao Xu, Haozhao Wang, Fushuo Huo, Jinyu Chen, Song Guo
Abstract: Selecting proper clients to participate in each federated learning (FL) round is critical to effectively harness a broad range of distributed data. Existing client selection methods simply consider the mining of distributed uni-modal data, yet, their effectiveness may diminish in multi-modal FL (MFL) as the modality imbalance problem not only impedes the collaborative local training but also leads to a severe global modality-level bias. We empirically reveal that local training with a certain single modality may contribute more to the global model than training with all local modalities. To effectively exploit the distributed multiple modalities, we propose a novel Balanced Modality Selection framework for MFL (BMSFed) to overcome the modal bias. On the one hand, we introduce a modal enhancement loss during local training to alleviate local imbalance based on the aggregated global prototypes. On the other hand, we propose the modality selection aiming to select subsets of local modalities with great diversity and achieving global modal balance simultaneously. Our extensive experiments on audio-visual, colored-gray, and front-back datasets showcase the superiority of BMSFed over baselines and its effectiveness in multi-modal data exploitation.
Authors: Chao Ma, Ziyang Wang
Abstract: Medical image segmentation is essential in diagnostics, treatment planning, and healthcare, with deep learning offering promising advancements. Notably, the convolutional neural network (CNN) excels in capturing local image features, whereas the Vision Transformer (ViT) adeptly models long-range dependencies through multi-head self-attention mechanisms. Despite their strengths, both the CNN and ViT face challenges in efficiently processing long-range dependencies in medical images, often requiring substantial computational resources. This issue, combined with the high cost and limited availability of expert annotations, poses significant obstacles to achieving precise segmentation. To address these challenges, this study introduces Semi-Mamba-UNet, which integrates a purely visual Mamba-based U-shaped encoder-decoder architecture with a conventional CNN-based UNet into a semi-supervised learning (SSL) framework. This innovative SSL approach leverages both networks to generate pseudo-labels and cross-supervise one another at the pixel level simultaneously, drawing inspiration from consistency regularisation techniques. Furthermore, we introduce a self-supervised pixel-level contrastive learning strategy that employs a pair of projectors to enhance the feature learning capabilities further, especially on unlabelled data. Semi-Mamba-UNet was comprehensively evaluated on two publicly available segmentation dataset and compared with seven other SSL frameworks with both CNN- or ViT-based UNet as the backbone network, highlighting the superior performance of the proposed method. The source code of Semi-Mamba-Unet, all baseline SSL frameworks, the CNN- and ViT-based networks, and the two corresponding datasets are made publicly accessible.
Authors: Teruyuki Katsuoka, Tomohiro Shiraishi, Daiki Miwa, Vo Nguyen Le Duy, Ichiro Takeuchi
Abstract: AI technology for generating images, such as diffusion models, has advanced rapidly. However, there is no established framework for quantifying the reliability of AI-generated images, which hinders their use in critical decision-making tasks, such as medical image diagnosis. In this study, we propose a method to quantify the reliability of decision-making tasks that rely on images produced by diffusion models within a statistical testing framework. The core concept of our statistical test involves using a selective inference framework, in which the statistical test is conducted under the condition that the images are produced by a trained diffusion model. As a case study, we study a diffusion model-based anomaly detection task for medical images. With our approach, the statistical significance of medical image diagnostic outcomes can be quantified in terms of a p-value, enabling decision-making with a controlled error rate. We demonstrate the theoretical soundness and practical effectiveness of our statistical test through numerical experiments on both synthetic and brain image datasets.
Authors: Minyoung Huh, Brian Cheung, Jeremy Bernstein, Phillip Isola, Pulkit Agrawal
Abstract: The scalability of deep learning models is fundamentally limited by computing resources, memory, and communication. Although methods like low-rank adaptation (LoRA) have reduced the cost of model finetuning, its application in model pre-training remains largely unexplored. This paper explores extending LoRA to model pre-training, identifying the inherent constraints and limitations of standard LoRA in this context. We introduce LoRA-the-Explorer (LTE), a novel bi-level optimization algorithm designed to enable parallel training of multiple low-rank heads across computing nodes, thereby reducing the need for frequent synchronization. Our approach includes extensive experimentation on vision transformers using various vision datasets, demonstrating that LTE is competitive with standard pre-training.
Authors: Zhiyuan Li, Yanhui Zhou, Hao Wei, Chenyang Ge, Jingwen Jiang
Abstract: Image compression at extremely low bitrates (below 0.1 bits per pixel (bpp)) is a significant challenge due to substantial information loss. In this work, we propose a novel two-stage extreme image compression framework that exploits the powerful generative capability of pre-trained diffusion models to achieve realistic image reconstruction at extremely low bitrates. In the first stage, we treat the latent representation of images in the diffusion space as guidance, employing a VAE-based compression approach to compress images and initially decode the compressed information into content variables. The second stage leverages pre-trained stable diffusion to reconstruct images under the guidance of content variables. Specifically, we introduce a small control module to inject content information while keeping the stable diffusion model fixed to maintain its generative capability. Furthermore, we design a space alignment loss to force the content variables to align with the diffusion space and provide the necessary constraints for optimization. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art approaches in terms of visual performance at extremely low bitrates.
Authors: Xinying Lin, Xuyang Liu, Hong Yang, Xiaohai He, Honggang Chen
Abstract: With the advent of image super-resolution (SR) algorithms, how to evaluate the quality of generated SR images has become an urgent task. Although full-reference methods perform well in SR image quality assessment (SR-IQA), their reliance on high-resolution (HR) images limits their practical applicability. Leveraging available reconstruction information as much as possible for SR-IQA, such as low-resolution (LR) images and the scale factors, is a promising way to enhance assessment performance for SR-IQA without HR for reference. In this letter, we attempt to evaluate the perceptual quality and reconstruction fidelity of SR images considering LR images and scale factors. Specifically, we propose a novel dual-branch reduced-reference SR-IQA network, \ie, Perception- and Fidelity-aware SR-IQA (PFIQA). The perception-aware branch evaluates the perceptual quality of SR images by leveraging the merits of global modeling of Vision Transformer (ViT) and local relation of ResNet, and incorporating the scale factor to enable comprehensive visual perception. Meanwhile, the fidelity-aware branch assesses the reconstruction fidelity between LR and SR images through their visual perception. The combination of the two branches substantially aligns with the human visual system, enabling a comprehensive SR image evaluation. Experimental results indicate that our PFIQA outperforms current state-of-the-art models across three widely-used SR-IQA benchmarks. Notably, PFIQA excels in assessing the quality of real-world SR images.
Authors: Xiyuan Wei, Fanjiang Ye, Ori Yonay, Xingyu Chen, Baixi Sun, Dingwen Tao, Tianbao Yang
Abstract: Existing studies of training state-of-the-art Contrastive Language-Image Pretraining (CLIP) models on large-scale data involve hundreds of or even thousands of GPUs due to the requirement of a large batch size. However, such a large amount of resources is not accessible to most people. While advanced compositional optimization techniques for optimizing global contrastive losses have been demonstrated effective for removing the requirement of large batch size, their performance on large-scale data remains underexplored and not optimized. To bridge the gap, this paper explores several aspects of CLIP training with limited resources (e.g., up to tens of GPUs). First, we introduce FastCLIP, a general CLIP training framework built on advanced compositional optimization techniques while designed and optimized for the distributed setting. Our framework is equipped with an efficient gradient reduction strategy to reduce communication overhead. Second, to further boost training efficiency, we investigate three components of the framework from an optimization perspective: the schedule of the inner learning rate, the update rules of the temperature parameter and the model parameters, respectively. Experiments on different strategies for each component shed light on how to conduct CLIP training more efficiently. Finally, we benchmark the performance of FastCLIP and the state-of-the-art training baseline (OpenCLIP) on different compute scales up to 32 GPUs on 8 nodes, and three data scales ranging from 2.7 million, 9.1 million to 315 million image-text pairs to demonstrate the significant improvement of FastCLIP in the resource-limited setting. We release the code of FastCLIP at https://github.com/Optimization-AI/fast_clip .
Authors: Fuqiang Chen, Ranran Zhang, Boyun Zheng, Yiwen Sun, Jiahui He, Wenjian Qin
Abstract: Conventional hematoxylin-eosin (H&E) staining is limited to revealing cell morphology and distribution, whereas immunohistochemical (IHC) staining provides precise and specific visualization of protein activation at the molecular level. Virtual staining technology has emerged as a solution for highly efficient IHC examination, which directly transforms H&E-stained images to IHC-stained images. However, virtual staining is challenged by the insufficient mining of pathological semantics and the spatial misalignment of pathological semantics. To address these issues, we propose the Pathological Semantics-Preserving Learning method for Virtual Staining (PSPStain), which directly incorporates the molecular-level semantic information and enhances semantics interaction despite any spatial inconsistency. Specifically, PSPStain comprises two novel learning strategies: 1) Protein-Aware Learning Strategy (PALS) with Focal Optical Density (FOD) map maintains the coherence of protein expression level, which represents molecular-level semantic information; 2) Prototype-Consistent Learning Strategy (PCLS), which enhances cross-image semantic interaction by prototypical consistency learning. We evaluate PSPStain on two public datasets using five metrics: three clinically relevant metrics and two for image quality. Extensive experiments indicate that PSPStain outperforms current state-of-the-art H&E-to-IHC virtual staining methods and demonstrates a high pathological correlation between the staging of real and virtual stains.
Authors: Arianna Bunnell, Kailee Hung, John A. Shepherd, Peter Sadowski
Abstract: Development of artificial intelligence (AI) for medical imaging demands curation and cleaning of large-scale clinical datasets comprising hundreds of thousands of images. Some modalities, such as mammography, contain highly standardized imaging. In contrast, breast ultrasound imaging (BUS) can contain many irregularities not indicated by scan metadata, such as enhanced scan modes, sonographer annotations, or additional views. We present an open-source software solution for automatically processing clinical BUS datasets. The algorithm performs BUS scan filtering, cleaning, and knowledge extraction from sonographer annotations. Its modular design enables users to adapt it to new settings. Experiments on an internal testing dataset of 430 clinical BUS images achieve >95% sensitivity and >98% specificity in detecting every type of text annotation, >98% sensitivity and specificity in detecting scans with blood flow highlighting, alternative scan modes, or invalid scans. A case study on a completely external, public dataset of BUS scans found that BUSClean identified text annotations and scans with blood flow highlighting with 88.6% and 90.9% sensitivity and 98.3% and 99.9% specificity, respectively. Adaptation of the lesion caliper detection method to account for a type of caliper specific to the case study demonstrates intended use of BUSClean in new data distributions and improved performance in lesion caliper detection from 43.3% and 93.3% out-of-the-box to 92.1% and 92.3% sensitivity and specificity, respectively. Source code, example notebooks, and sample data are available at https://github.com/hawaii-ai/bus-cleaning.
Authors: Jixiang Luo
Abstract: Since LIC has made rapid progress recently compared to traditional methods, this paper attempts to discuss the question about 'Where is the boundary of Learned Image Compression(LIC)?' with regard to subjective matrics. Thus this paper splits the above problem into two sub-problems:1)Where is the boundary of rate-distortion performance of PSNR? 2)How to further improve the compression gain and achieve the boundary? Therefore this paper analyzes the effectiveness of scaling parameters for encoder, decoder and context model, which are the three components of LIC. Then we conclude that scaling for LIC is to scale for context model and decoder within LIC. Extensive experiments demonstrate that overfitting can actually serve as an effective context. By optimizing the context, this paper further improves PSNR and achieves state-of-the-art performance, showing a performance gain of 14.39% with BD-RATE over VVC.
Authors: Zeyu Wang, Weichen Dai, Xiangyu Zhou, Ji Qi, Yi Zhou
Abstract: Vision Transformer and its variants have been adopted in many visual tasks due to their powerful capabilities, which also bring significant challenges in computation and storage. Consequently, researchers have introduced various compression methods in recent years, among which the pruning techniques are widely used to remove a significant fraction of the network. Therefore, these methods can reduce significant percent of the FLOPs, but often lead to a decrease in model performance. To investigate the underlying causes, we focus on the pruning methods specifically belonging to the pruning-during-training category, then drew inspiration from neuroscience and propose a new concept for artificial neural network models named Neural Burden. We investigate its impact in the model pruning process, and subsequently explore a simple yet effective approach to mitigate the decline in model performance, which can be applied to any pruning-during-training technique. Extensive experiments indicate that the neural burden phenomenon indeed exists, and show the potential of our method. We hope that our findings can provide valuable insights for future research. Code will be made publicly available after this paper is published.