Authors: Dominykas Seputis, Serghei Mihailov, Soham Chatterjee, Zehao Xiao
Abstract: Large pre-trained vision-language models, such as CLIP, have demonstrated state-of-the-art performance across a wide range of image classification tasks, without requiring retraining. Few-shot CLIP is competitive with existing specialized architectures that were trained on the downstream tasks. Recent research demonstrates that the performance of CLIP can be further improved using lightweight adaptation approaches. However, previous methods adapt different modalities of the CLIP model individually, ignoring the interactions and relationships between visual and textual representations. In this work, we propose Multi-Modal Adapter, an approach for Multi-Modal adaptation of CLIP. Specifically, we add a trainable Multi-Head Attention layer that combines text and image features to produce an additive adaptation of both. Multi-Modal Adapter demonstrates improved generalizability, based on its performance on unseen classes compared to existing adaptation methods. We perform additional ablations and investigations to validate and interpret the proposed approach.
Authors: Haiyu Wu, Jaskirat Singh, Sicong Tian, Liang Zheng, Kevin W. Bowyer
Abstract: This paper studies how to synthesize face images of non-existent persons, to create a dataset that allows effective training of face recognition (FR) models. Two important goals are (1) the ability to generate a large number of distinct identities (inter-class separation) with (2) a wide variation in appearance of each identity (intra-class variation). However, existing works 1) are typically limited in how many well-separated identities can be generated and 2) either neglect or use a separate editing model for attribute augmentation. We propose Vec2Face, a holistic model that uses only a sampled vector as input and can flexibly generate and control face images and their attributes. Composed of a feature masked autoencoder and a decoder, Vec2Face is supervised by face image reconstruction and can be conveniently used in inference. Using vectors with low similarity among themselves as inputs, Vec2Face generates well-separated identities. Randomly perturbing an input identity vector within a small range allows Vec2Face to generate faces of the same identity with robust variation in face attributes. It is also possible to generate images with designated attributes by adjusting vector values with a gradient descent method. Vec2Face has efficiently synthesized as many as 300K identities with 15 million total images, whereas 60K is the largest number of identities created in the previous works. FR models trained with the generated HSFace datasets, from 10k to 300k identities, achieve state-of-the-art accuracy, from 92% to 93.52%, on five real-world test sets. For the first time, our model created using a synthetic training set achieves higher accuracy than the model created using a same-scale training set of real face images (on the CALFW test set).
Authors: Mehmet Kerem Turkcan, Ian Li, Chengbo Zang, Javad Ghaderi, Gil Zussman, Zoran Kostic
Abstract: We introduce Boundless, a photo-realistic synthetic data generation system for enabling highly accurate object detection in dense urban streetscapes. Boundless can replace massive real-world data collection and manual ground-truth object annotation (labeling) with an automated and configurable process. Boundless is based on the Unreal Engine 5 (UE5) City Sample project with improvements enabling accurate collection of 3D bounding boxes across different lighting and scene variability conditions. We evaluate the performance of object detection models trained on the dataset generated by Boundless when used for inference on a real-world dataset acquired from medium-altitude cameras. We compare the performance of the Boundless-trained model against the CARLA-trained model and observe an improvement of 7.8 mAP. The results we achieved support the premise that synthetic data generation is a credible methodology for training/fine-tuning scalable object detection models for urban scenes.
Authors: Manu Gaur, Darshan Singh S, Makarand Tapaswi
Abstract: Image captioning systems are unable to generate fine-grained captions as they are trained on data that is either noisy (alt-text) or generic (human annotations). This is further exacerbated by maximum likelihood training that encourages generation of frequently occurring phrases. Previous works have tried to address this limitation by fine-tuning captioners with a self-retrieval (SR) reward. However, we find that SR fine-tuning has a tendency to reduce caption faithfulness and even hallucinate. In this work, we circumvent this bottleneck by improving the MLE initialization of the captioning system and designing a curriculum for the SR fine-tuning process. To this extent, we present (1) Visual Caption Boosting, a novel framework to instill fine-grainedness in generic image captioning datasets while remaining anchored in human annotations; and (2) BagCurri, a carefully designed training curriculum that more optimally leverages the contrastive nature of the self-retrieval reward. Jointly, they enable the captioner to describe fine-grained aspects in the image while preserving faithfulness to ground-truth captions. Our approach outperforms previous work by +8.9% on SR against 99 random distractors (RD100) (Dessi et al., 2023); and +7.6% on ImageCoDe. Additionally, existing metrics to evaluate captioning systems fail to reward diversity or evaluate a model's fine-grained understanding ability. Our third contribution addresses this by proposing self-retrieval from the lens of evaluation. We introduce TrueMatch, a benchmark comprising bags of highly similar images that uses SR to assess the captioner's ability to capture subtle visual distinctions. We evaluate and compare several state-of-the-art open-source MLLMs on TrueMatch, and find that our SR approach outperforms them all by a significant margin (e.g. +4.8% - 7.1% over Cambrian) while having 1-2 orders of magnitude fewer parameters.
Authors: Shuang Song, Rongjun Qin
Abstract: Modeling outdoor scenes for the synthetic 3D environment requires the recovery of reflectance/albedo information from raw images, which is an ill-posed problem due to the complicated unmodeled physics in this process (e.g., indirect lighting, volume scattering, specular reflection). The problem remains unsolved in a practical context. The recovered albedo can facilitate model relighting and shading, which can further enhance the realism of rendered models and the applications of digital twins. Typically, photogrammetric 3D models simply take the source images as texture materials, which inherently embed unwanted lighting artifacts (at the time of capture) into the texture. Therefore, these polluted textures are suboptimal for a synthetic environment to enable realistic rendering. In addition, these embedded environmental lightings further bring challenges to photo-consistencies across different images that cause image-matching uncertainties. This paper presents a general image formation model for albedo recovery from typical aerial photogrammetric images under natural illuminations and derives the inverse model to resolve the albedo information through inverse rendering intrinsic image decomposition. Our approach builds on the fact that both the sun illumination and scene geometry are estimable in aerial photogrammetry, thus they can provide direct inputs for this ill-posed problem. This physics-based approach does not require additional input other than data acquired through the typical drone-based photogrammetric collection and was shown to favorably outperform existing approaches. We also demonstrate that the recovered albedo image can in turn improve typical image processing tasks in photogrammetry such as feature and dense matching, edge, and line extraction.
Authors: Avigail Cohen Rimon, Tal Shnitzer, Mirela Ben Chen
Abstract: We propose a novel framework for representing neural fields on triangle meshes that is multi-resolution across both spatial and frequency domains. Inspired by the Neural Fourier Filter Bank (NFFB), our architecture decomposes the spatial and frequency domains by associating finer spatial resolution levels with higher frequency bands, while coarser resolutions are mapped to lower frequencies. To achieve geometry-aware spatial decomposition we leverage multiple DiffusionNet components, each associated with a different spatial resolution level. Subsequently, we apply a Fourier feature mapping to encourage finer resolution levels to be associated with higher frequencies. The final signal is composed in a wavelet-inspired manner using a sine-activated MLP, aggregating higher-frequency signals on top of lower-frequency ones. Our architecture attains high accuracy in learning complex neural fields and is robust to discontinuities, exponential scale variations of the target field, and mesh modification. We demonstrate the effectiveness of our approach through its application to diverse neural fields, such as synthetic RGB functions, UV texture coordinates, and vertex normals, illustrating different challenges. To validate our method, we compare its performance against two alternatives, showcasing the advantages of our multi-resolution architecture.
Authors: Christiaan Viviers, Amaan Valiuddin, Francisco Caetano, Lemar Abdi, Lena Filatova, Peter de With, Fons van der Sommen
Abstract: Detecting Out-of-Distribution~(OOD) sensory data and covariate distribution shift aims to identify new test examples with different high-level image statistics to the captured, normal and In-Distribution (ID) set. Existing OOD detection literature largely focuses on semantic shift with little-to-no consensus over covariate shift. Generative models capture the ID data in an unsupervised manner, enabling them to effectively identify samples that deviate significantly from this learned distribution, irrespective of the downstream task. In this work, we elucidate the ability of generative models to detect and quantify domain-specific covariate shift through extensive analyses that involves a variety of models. To this end, we conjecture that it is sufficient to detect most occurring sensory faults (anomalies and deviations in global signals statistics) by solely modeling high-frequency signal-dependent and independent details. We propose a novel method, CovariateFlow, for OOD detection, specifically tailored to covariate heteroscedastic high-frequency image-components using conditional Normalizing Flows (cNFs). Our results on CIFAR10 vs. CIFAR10-C and ImageNet200 vs. ImageNet200-C demonstrate the effectiveness of the method by accurately detecting OOD covariate shift. This work contributes to enhancing the fidelity of imaging systems and aiding machine learning models in OOD detection in the presence of covariate shift.
Authors: Arkadeep Narayan Chaudhury, Igor Vasiljevic, Sergey Zakharov, Vitor Guizilini, Rares Ambrus, Srinivasa Narasimhan, Christopher G. Atkeson
Abstract: Synthesizing accurate geometry and photo-realistic appearance of small scenes is an active area of research with compelling use cases in gaming, virtual reality, robotic-manipulation, autonomous driving, convenient product capture, and consumer-level photography. When applying scene geometry and appearance estimation techniques to robotics, we found that the narrow cone of possible viewpoints due to the limited range of robot motion and scene clutter caused current estimation techniques to produce poor quality estimates or even fail. On the other hand, in robotic applications, dense metric depth can often be measured directly using stereo and illumination can be controlled. Depth can provide a good initial estimate of the object geometry to improve reconstruction, while multi-illumination images can facilitate relighting. In this work we demonstrate a method to incorporate dense metric depth into the training of neural 3D representations and address an artifact observed while jointly refining geometry and appearance by disambiguating between texture and geometry edges. We also discuss a multi-flash stereo camera system developed to capture the necessary data for our pipeline and show results on relighting and view synthesis with a few training views.
Authors: Shehan Perera, Yunus Erzurumlu, Deepak Gulati, Alper Yilmaz
Abstract: Skin cancer segmentation poses a significant challenge in medical image analysis. Numerous existing solutions, predominantly CNN-based, face issues related to a lack of global contextual understanding. Alternatively, some approaches resort to large-scale Transformer models to bridge the global contextual gaps, but at the expense of model size and computational complexity. Finally many Transformer based approaches rely primarily on CNN based decoders overlooking the benefits of Transformer based decoding models. Recognizing these limitations, we address the need efficient lightweight solutions by introducing MobileUNETR, which aims to overcome the performance constraints associated with both CNNs and Transformers while minimizing model size, presenting a promising stride towards efficient image segmentation. MobileUNETR has 3 main features. 1) MobileUNETR comprises of a lightweight hybrid CNN-Transformer encoder to help balance local and global contextual feature extraction in an efficient manner; 2) A novel hybrid decoder that simultaneously utilizes low-level and global features at different resolutions within the decoding stage for accurate mask generation; 3) surpassing large and complex architectures, MobileUNETR achieves superior performance with 3 million parameters and a computational complexity of 1.3 GFLOP resulting in 10x and 23x reduction in parameters and FLOPS, respectively. Extensive experiments have been conducted to validate the effectiveness of our proposed method on four publicly available skin lesion segmentation datasets, including ISIC 2016, ISIC 2017, ISIC 2018, and PH2 datasets. The code will be publicly available at: https://github.com/OSUPCVLab/MobileUNETR.git
Authors: Chen Li, Xiaoling Hu, Shahira Abousamra, Meilong Xu, Chao Chen
Abstract: Generative models, such as GANs and diffusion models, have been used to augment training sets and boost performances in different tasks. We focus on generative models for cell detection instead, i.e., locating and classifying cells in given pathology images. One important information that has been largely overlooked is the spatial patterns of the cells. In this paper, we propose a spatial-pattern-guided generative model for cell layout generation. Specifically, a novel diffusion model guided by spatial features and generates realistic cell layouts has been proposed. We explore different density models as spatial features for the diffusion model. In downstream tasks, we show that the generated cell layouts can be used to guide the generation of high-quality pathology images. Augmenting with these images can significantly boost the performance of SOTA cell detection methods. The code is available at https://github.com/superlc1995/Diffusion-cell.
Authors: Mamadou Keita, Wassim Hamidouche, Hessen Bougueffa Eutamene, Abdelmalik Taleb-Ahmed, Abdenour Hadid
Abstract: We introduce FIDAVL: Fake Image Detection and Attribution using a Vision-Language Model. FIDAVL is a novel and efficient mul-titask approach inspired by the synergies between vision and language processing. Leveraging the benefits of zero-shot learning, FIDAVL exploits the complementarity between vision and language along with soft prompt-tuning strategy to detect fake images and accurately attribute them to their originating source models. We conducted extensive experiments on a comprehensive dataset comprising synthetic images generated by various state-of-the-art models. Our results demonstrate that FIDAVL achieves an encouraging average detection accuracy of 95.42% and F1-score of 95.47% while also obtaining noteworthy performance metrics, with an average F1-score of 92.64% and ROUGE-L score of 96.50% for attributing synthetic images to their respective source generation models. The source code of this work will be publicly released at https://github.com/Mamadou-Keita/FIDAVL.
Authors: Yike Zhang, Jack Noble
Abstract: Cochlear Implant (CI) procedures involve performing an invasive mastoidectomy to insert an electrode array into the cochlea. In this paper, we introduce a novel pipeline that is capable of generating synthetic multi-view videos from a single CI microscope image. In our approach, we use a patient's pre-operative CT scan to predict the post-mastoidectomy surface using a method designed for this purpose. We manually align the surface with a selected microscope frame to obtain an accurate initial pose of the reconstructed CT mesh relative to the microscope. We then perform UV projection to transfer the colors from the frame to surface textures. Novel views of the textured surface can be used to generate a large dataset of synthetic frames with ground truth poses. We evaluated the quality of synthetic views rendered using Pytorch3D and PyVista. We found both rendering engines lead to similarly high-quality synthetic novel-view frames compared to ground truth with a structural similarity index for both methods averaging about 0.86. A large dataset of novel views with known poses is critical for ongoing training of a method to automatically estimate microscope pose for 2D to 3D registration with the pre-operative CT to facilitate augmented reality surgery. This dataset will empower various downstream tasks, such as integrating Augmented Reality (AR) in the OR, tracking surgical tools, and supporting other video analysis studies.
Authors: Bowen Tian, Songning Lai, Lujundong Li, Zhihao Shuai, Runwei Guan, Tian Wu, Yutao Yue
Abstract: Fine-grained image classification has witnessed significant advancements with the advent of deep learning and computer vision technologies. However, the scarcity of detailed annotations remains a major challenge, especially in scenarios where obtaining high-quality labeled data is costly or time-consuming. To address this limitation, we introduce Precision-Enhanced Pseudo-Labeling(PEPL) approach specifically designed for fine-grained image classification within a semi-supervised learning framework. Our method leverages the abundance of unlabeled data by generating high-quality pseudo-labels that are progressively refined through two key phases: initial pseudo-label generation and semantic-mixed pseudo-label generation. These phases utilize Class Activation Maps (CAMs) to accurately estimate the semantic content and generate refined labels that capture the essential details necessary for fine-grained classification. By focusing on semantic-level information, our approach effectively addresses the limitations of standard data augmentation and image-mixing techniques in preserving critical fine-grained features. We achieve state-of-the-art performance on benchmark datasets, demonstrating significant improvements over existing semi-supervised strategies, with notable boosts in accuracy and robustness.Our code has been open sourced at https://github.com/TianSuya/SemiFG.
Authors: Zhaowei Wang, Ying Hao, Hao Wei, Qing Xiao, Lulu Chen, Yulong Li, Yue Yang, Tianyi Li
Abstract: Recent advancements in text-to-image diffusion models have significantly transformed visual content generation, yet their application in specialized fields such as interior design remains underexplored. In this paper, we present RoomDiffusion, a pioneering diffusion model meticulously tailored for the interior design industry. To begin with, we build from scratch a whole data pipeline to update and evaluate data for iterative model optimization. Subsequently, techniques such as multiaspect training, multi-stage fine-tune and model fusion are applied to enhance both the visual appeal and precision of the generated results. Lastly, leveraging the latent consistency Distillation method, we distill and expedite the model for optimal efficiency. Unlike existing models optimized for general scenarios, RoomDiffusion addresses specific challenges in interior design, such as lack of fashion, high furniture duplication rate, and inaccurate style. Through our holistic human evaluation protocol with more than 20 professional human evaluators, RoomDiffusion demonstrates industry-leading performance in terms of aesthetics, accuracy, and efficiency, surpassing all existing open source models such as stable diffusion and SDXL.
Authors: Pu Sun, Honggang Qi, Yuezun Li
Abstract: DeepFake technology has gained significant attention due to its ability to manipulate facial attributes with high realism, raising serious societal concerns. Face-Swap DeepFake is the most harmful among these techniques, which fabricates behaviors by swapping original faces with synthesized ones. Existing forensic methods, primarily based on Deep Neural Networks (DNNs), effectively expose these manipulations and have become important authenticity indicators. However, these methods mainly concentrate on capturing the blending inconsistency in DeepFake faces, raising a new security issue, termed Active Fake, emerges when individuals intentionally create blending inconsistency in their authentic videos to evade responsibility. This tactic is called DeepFake Camouflage. To achieve this, we introduce a new framework for creating DeepFake camouflage that generates blending inconsistencies while ensuring imperceptibility, effectiveness, and transferability. This framework, optimized via an adversarial learning strategy, crafts imperceptible yet effective inconsistencies to mislead forensic detectors. Extensive experiments demonstrate the effectiveness and robustness of our method, highlighting the need for further research in active fake detection.
Authors: Mingze Gao, Jingyu Liu, Mingda Li, Jiangtao Xie, Qingbin Liu, Bo Zhao, Xi Chen, Hui Xiong
Abstract: Multimodal Large Language Models (MLLMs) have significantly improved performance across various image-language applications. Recently, there has been a growing interest in adapting image pre-trained MLLMs for video-related tasks. However, most efforts concentrate on enhancing the vision encoder and projector components, while the core part, Large Language Models (LLMs), remains comparatively under-explored. In this paper, we propose two strategies to enhance the model's capability in video understanding tasks by improving inter-layer attention computation in LLMs. Specifically, the first approach focuses on the enhancement of Rotary Position Embedding (RoPE) with Temporal-Aware Dual RoPE, which introduces temporal position information to strengthen the MLLM's temporal modeling capabilities while preserving the relative position relationships of both visual and text tokens. The second approach involves enhancing the Attention Mask with the Frame-wise Block Causal Attention Mask, a simple yet effective method that broadens visual token interactions within and across video frames while maintaining the causal inference mechanism. Based on these proposed methods, we adapt LLaVA for video understanding tasks, naming it Temporal-Considered LLaVA (TC-LLaVA). Our TC-LLaVA achieves new state-of-the-art performance across various video understanding benchmarks with only supervised fine-tuning (SFT) on video-related datasets.
Authors: Lin Sun, Jiale Cao, Jin Xie, Fahad Shahbaz Khan, Yanwei Pang
Abstract: Stable diffusion has demonstrated strong image synthesis ability to given text descriptions, suggesting it to contain strong semantic clue for grouping objects. Inspired by this, researchers have explored employing stable diffusion for trainingfree segmentation. Most existing approaches either simply employ cross-attention map or refine it by self-attention map, to generate segmentation masks. We believe that iterative refinement with self-attention map would lead to better results. However, we mpirically demonstrate that such a refinement is sub-optimal likely due to the self-attention map containing irrelevant global information which hampers accurately refining cross-attention map with multiple iterations. To address this, we propose an iterative refinement framework for training-free segmentation, named iSeg, having an entropy-reduced self-attention module which utilizes a gradient descent scheme to reduce the entropy of self-attention map, thereby suppressing the weak responses corresponding to irrelevant global information. Leveraging the entropy-reduced self-attention module, our iSeg stably improves refined crossattention map with iterative refinement. Further, we design a category-enhanced cross-attention module to generate accurate cross-attention map, providing a better initial input for iterative refinement. Extensive experiments across different datasets and diverse segmentation tasks reveal the merits of proposed contributions, leading to promising performance on diverse segmentation tasks. For unsupervised semantic segmentation on Cityscapes, our iSeg achieves an absolute gain of 3.8% in terms of mIoU compared to the best existing training-free approach in literature. Moreover, our proposed iSeg can support segmentation with different kind of images and interactions.
Authors: Hersh Vakharia, Xiaoxiao Du
Abstract: Sensor fusion combines data from multiple sensor sources to improve reliability, robustness, and accuracy of data interpretation. The Fuzzy Integral (FI), in particular, the Choquet integral (ChI), is often used as a powerful nonlinear aggregator for fusion across multiple sensors. However, existing supervised ChI learning algorithms typically require precise training labels for each input data point, which can be difficult or impossible to obtain. Additionally, prior work on ChI fusion is often based only on the normalized fuzzy measures, which bounds the fuzzy measure values between [0, 1]. This can be limiting in cases where the underlying scales of input data sources are bipolar (i.e., between [-1, 1]). To address these challenges, this paper proposes a novel Choquet integral-based fusion framework, named Bi-MIChI (pronounced "bi-mi-kee"), which uses bi-capacities to represent the interactions between pairs of subsets of the input sensor sources on a bi-polar scale. This allows for extended non-linear interactions between the sensor sources and can lead to interesting fusion results. Bi-MIChI also addresses label uncertainty through Multiple Instance Learning, where training labels are applied to "bags" (sets) of data instead of per-instance. Our proposed Bi-MIChI framework shows effective classification and detection performance on both synthetic and real-world experiments for sensor fusion with label uncertainty. We also provide detailed analyses on the behavior of the fuzzy measures to demonstrate our fusion process.
Authors: Shen Chen, Jiale Zhou, Lei Li
Abstract: 3D Gaussian Splatting (3DGS) has emerged as a promising approach for 3D scene representation, offering a reduction in computational overhead compared to Neural Radiance Fields (NeRF). However, 3DGS is susceptible to high-frequency artifacts and demonstrates suboptimal performance under sparse viewpoint conditions, thereby limiting its applicability in robotics and computer vision. To address these limitations, we introduce SVS-GS, a novel framework for Sparse Viewpoint Scene reconstruction that integrates a 3D Gaussian smoothing filter to suppress artifacts. Furthermore, our approach incorporates a Depth Gradient Profile Prior (DGPP) loss with a dynamic depth mask to sharpen edges and 2D diffusion with Score Distillation Sampling (SDS) loss to enhance geometric consistency in novel view synthesis. Experimental evaluations on the MipNeRF-360 and SeaThru-NeRF datasets demonstrate that SVS-GS markedly improves 3D reconstruction from sparse viewpoints, offering a robust and efficient solution for scene understanding in robotics and computer vision applications.
Authors: Chenguang Zhu, Shan Gao, Huafeng Chen, Guangqian Guo, Chaowei Wang, Yaoxing Wang, Chen Shu Lei, Quanjiang Fan
Abstract: Multi-modality image fusion aims to integrate the merits of images from different sources and render high-quality fusion images. However, existing feature extraction and fusion methods are either constrained by inherent local reduction bias and static parameters during inference (CNN) or limited by quadratic computational complexity (Transformers), and cannot effectively extract and fuse features. To solve this problem, we propose a dual-branch image fusion network called Tmamba. It consists of linear Transformer and Mamba, which has global modeling capabilities while maintaining linear complexity. Due to the difference between the Transformer and Mamba structures, the features extracted by the two branches carry channel and position information respectively. T-M interaction structure is designed between the two branches, using global learnable parameters and convolutional layers to transfer position and channel information respectively. We further propose cross-modal interaction at the attention level to obtain cross-modal attention. Experiments show that our Tmamba achieves promising results in multiple fusion tasks, including infrared-visible image fusion and medical image fusion. Code with checkpoints will be available after the peer-review process.
Authors: Xixi Jiang, Dong Zhang, Xiang Li, Kangyi Liu, Kwang-Ting Cheng, Xin Yang
Abstract: Partially-supervised multi-organ medical image segmentation aims to develop a unified semantic segmentation model by utilizing multiple partially-labeled datasets, with each dataset providing labels for a single class of organs. However, the limited availability of labeled foreground organs and the absence of supervision to distinguish unlabeled foreground organs from the background pose a significant challenge, which leads to a distribution mismatch between labeled and unlabeled pixels. Although existing pseudo-labeling methods can be employed to learn from both labeled and unlabeled pixels, they are prone to performance degradation in this task, as they rely on the assumption that labeled and unlabeled pixels have the same distribution. In this paper, to address the problem of distribution mismatch, we propose a labeled-to-unlabeled distribution alignment (LTUDA) framework that aligns feature distributions and enhances discriminative capability. Specifically, we introduce a cross-set data augmentation strategy, which performs region-level mixing between labeled and unlabeled organs to reduce distribution discrepancy and enrich the training set. Besides, we propose a prototype-based distribution alignment method that implicitly reduces intra-class variation and increases the separation between the unlabeled foreground and background. This can be achieved by encouraging consistency between the outputs of two prototype classifiers and a linear classifier. Extensive experimental results on the AbdomenCT-1K dataset and a union of four benchmark datasets (including LiTS, MSD-Spleen, KiTS, and NIH82) demonstrate that our method outperforms the state-of-the-art partially-supervised methods by a considerable margin, and even surpasses the fully-supervised methods. The source code is publicly available at https://github.com/xjiangmed/LTUDA.
Authors: Chenglizhao Chen, Xinyu Liu, Mengke Song, Luming Li, Xu Yu, Shanchen Pang
Abstract: Detecting anomalies in human-related videos is crucial for surveillance applications. Current methods primarily include appearance-based and action-based techniques. Appearance-based methods rely on low-level visual features such as color, texture, and shape. They learn a large number of pixel patterns and features related to known scenes during training, making them effective in detecting anomalies within these familiar contexts. However, when encountering new or significantly changed scenes, i.e., unknown scenes, they often fail because existing SOTA methods do not effectively capture the relationship between actions and their surrounding scenes, resulting in low generalization. In contrast, action-based methods focus on detecting anomalies in human actions but are usually less informative because they tend to overlook the relationship between actions and their scenes, leading to incorrect detection. For instance, the normal event of running on the beach and the abnormal event of running on the street might both be considered normal due to the lack of scene information. In short, current methods struggle to integrate low-level visual and high-level action features, leading to poor anomaly detection in varied and complex scenes. To address this challenge, we propose a novel decoupling-based architecture for human-related video anomaly detection (DecoAD). DecoAD significantly improves the integration of visual and action features through the decoupling and interweaving of scenes and actions, thereby enabling a more intuitive and accurate understanding of complex behaviors and scenes. DecoAD supports fully supervised, weakly supervised, and unsupervised settings.
Authors: Md. Mahfuzur Rahman, Sunzida Siddique, Marufa Kamal, Rakib Hossain Rifat, Kishor Datta Gupta
Abstract: Unmanned Aerial Vehicles (UAVs), have greatly revolutionized the process of gathering and analyzing data in diverse research domains, providing unmatched adaptability and effectiveness. This paper presents a thorough examination of Unmanned Aerial Vehicle (UAV) datasets, emphasizing their wide range of applications and progress. UAV datasets consist of various types of data, such as satellite imagery, images captured by drones, and videos. These datasets can be categorized as either unimodal or multimodal, offering a wide range of detailed and comprehensive information. These datasets play a crucial role in disaster damage assessment, aerial surveillance, object recognition, and tracking. They facilitate the development of sophisticated models for tasks like semantic segmentation, pose estimation, vehicle re-identification, and gesture recognition. By leveraging UAV datasets, researchers can significantly enhance the capabilities of computer vision models, thereby advancing technology and improving our understanding of complex, dynamic environments from an aerial perspective. This review aims to encapsulate the multifaceted utility of UAV datasets, emphasizing their pivotal role in driving innovation and practical applications in multiple domains.
Authors: Yang Wen, Anyu Lai, Bo Qian, Hao Wang, Wuzhen Shi, Wenming Cao
Abstract: The current state-of-the-art in severe weather removal predominantly focuses on single-task applications, such as rain removal, haze removal, and snow removal. However, real-world weather conditions often consist of a mixture of several weather types, and the degree of weather mixing in autonomous driving scenarios remains unknown. In the presence of complex and diverse weather conditions, a single weather removal model often encounters challenges in producing clear images from severe weather images. Therefore, there is a need for the development of multi-task severe weather removal models that can effectively handle mixed weather conditions and improve image quality in autonomous driving scenarios. In this paper, we introduce a novel multi-task severe weather removal model that can effectively handle complex weather conditions in an adaptive manner. Our model incorporates a weather task sequence generator, enabling the self-attention mechanism to selectively focus on features specific to different weather types. To tackle the challenge of repairing large areas of weather degradation, we introduce Fast Fourier Convolution (FFC) to increase the receptive field. Additionally, we propose an adaptive upsampling technique that effectively processes both the weather task information and underlying image features by selectively retaining relevant information. Our proposed model has achieved state-of-the-art performance on the publicly available dataset.
Authors: Keisuke Toida, Naoki Kato, Osamu Segawa, Takeshi Nakamura, Kazuhiro Hotta
Abstract: We propose a Ground IoU (Gr-IoU) to address the data association problem in multi-object tracking. When tracking objects detected by a camera, it often occurs that the same object is assigned different IDs in consecutive frames, especially when objects are close to each other or overlapping. To address this issue, we introduce Gr-IoU, which takes into account the 3D structure of the scene. Gr-IoU transforms traditional bounding boxes from the image space to the ground plane using the vanishing point geometry. The IoU calculated with these transformed bounding boxes is more sensitive to the front-to-back relationships of objects, thereby improving data association accuracy and reducing ID switches. We evaluated our Gr-IoU method on the MOT17 and MOT20 datasets, which contain diverse tracking scenarios including crowded scenes and sequences with frequent occlusions. Experimental results demonstrated that Gr-IoU outperforms conventional real-time methods without appearance features.
Authors: Dawei Dai, Hao Zhu, Shuyin Xia, Guoyin Wang
Abstract: In actual scenarios, whether manually or automatically annotated, label noise is inevitably generated in the training data, which can affect the effectiveness of deep CNN models. The popular solutions require data cleaning or designing additional optimizations to punish the data with mislabeled data, thereby enhancing the robustness of models. However, these methods come at the cost of weakening or even losing some data during the training process. As we know, content is the inherent attribute of an image that does not change with changes in annotations. In this study, we propose a general granular-ball computing (GBC) module that can be embedded into a CNN model, where the classifier finally predicts the label of granular-ball ($gb$) samples instead of each individual samples. Specifically, considering the classification task: (1) in forward process, we split the input samples as $gb$ samples at feature-level, each of which can correspond to multiple samples with varying numbers and share one single label; (2) during the backpropagation process, we modify the gradient allocation strategy of the GBC module to enable it to propagate normally; and (3) we develop an experience replay policy to ensure the stability of the training process. Experiments demonstrate that the proposed method can improve the robustness of CNN models with no additional data or optimization.
Authors: Jinhee Kim, Taesung Kim, Jaegul Choo
Abstract: Recent advances in interactive keypoint estimation methods have enhanced accuracy while minimizing user intervention. However, these methods require user input for error correction, which can be costly in vertebrae keypoint estimation where inaccurate keypoints are densely clustered or overlap. We introduce a novel approach, KeyBot, specifically designed to identify and correct significant and typical errors in existing models, akin to user revision. By characterizing typical error types and using simulated errors for training, KeyBot effectively corrects these errors and significantly reduces user workload. Comprehensive quantitative and qualitative evaluations on three public datasets confirm that KeyBot significantly outperforms existing methods, achieving state-of-the-art performance in interactive vertebrae keypoint estimation. The source code and demo video are available at: https://ts-kim.github.io/KeyBot/
Authors: Weipeng Tan, Chuming Lin, Chengming Xu, Xiaozhong Ji, Junwei Zhu, Chengjie Wang, Yanwei Fu
Abstract: Talking Head Generation (THG), typically driven by audio, is an important and challenging task with broad application prospects in various fields such as digital humans, film production, and virtual reality. While diffusion model-based THG methods present high quality and stable content generation, they often overlook the intrinsic style which encompasses personalized features such as speaking habits and facial expressions of a video. As consequence, the generated video content lacks diversity and vividness, thus being limited in real life scenarios. To address these issues, we propose a novel framework named Style-Enhanced Vivid Portrait (SVP) which fully leverages style-related information in THG. Specifically, we first introduce the novel probabilistic style prior learning to model the intrinsic style as a Gaussian distribution using facial expressions and audio embedding. The distribution is learned through the 'bespoked' contrastive objective, effectively capturing the dynamic style information in each video. Then we finetune a pretrained Stable Diffusion (SD) model to inject the learned intrinsic style as a controlling signal via cross attention. Experiments show that our model generates diverse, vivid, and high-quality videos with flexible control over intrinsic styles, outperforming existing state-of-the-art methods.
Authors: Julong Wei, Shanshuai Yuan, Pengfei Li, Qingda Hu, Zhongxue Gan, Wenchao Ding
Abstract: The rise of multi-modal large language models(MLLMs) has spurred their applications in autonomous driving. Recent MLLM-based methods perform action by learning a direct mapping from perception to action, neglecting the dynamics of the world and the relations between action and world dynamics. In contrast, human beings possess world model that enables them to simulate the future states based on 3D internal visual representation and plan actions accordingly. To this end, we propose OccLLaMA, an occupancy-language-action generative world model, which uses semantic occupancy as a general visual representation and unifies vision-language-action(VLA) modalities through an autoregressive model. Specifically, we introduce a novel VQVAE-like scene tokenizer to efficiently discretize and reconstruct semantic occupancy scenes, considering its sparsity and classes imbalance. Then, we build a unified multi-modal vocabulary for vision, language and action. Furthermore, we enhance LLM, specifically LLaMA, to perform the next token/scene prediction on the unified vocabulary to complete multiple tasks in autonomous driving. Extensive experiments demonstrate that OccLLaMA achieves competitive performance across multiple tasks, including 4D occupancy forecasting, motion planning, and visual question answering, showcasing its potential as a foundation model in autonomous driving.
Authors: Jingyu Zhang, Wenqing Zhang, Chaoyi Tan, Xiangtian Li, Qianyi Sun
Abstract: It is very important to detect traffic signs efficiently and accurately in autonomous driving systems. However, the farther the distance, the smaller the traffic signs. Existing object detection algorithms can hardly detect these small scaled signs.In addition, the performance of embedded devices on vehicles limits the scale of detection models.To address these challenges, a YOLO PPA based traffic sign detection algorithm is proposed in this paper.The experimental results on the GTSDB dataset show that compared to the original YOLO, the proposed method improves inference efficiency by 11.2%. The mAP 50 is also improved by 93.2%, which demonstrates the effectiveness of the proposed YOLO PPA.
Authors: Huaxi Huang, Xin Yuan, Qiyu Liao, Dadong Wang, Tongliang Liu
Abstract: In the realm of multimedia data analysis, the extensive use of image datasets has escalated concerns over privacy protection within such data. Current research predominantly focuses on privacy protection either in data sharing or upon the release of trained machine learning models. Our study pioneers a comprehensive privacy protection framework that safeguards image data privacy concurrently during data sharing and model publication. We propose an interactive image privacy protection framework that utilizes generative machine learning models to modify image information at the attribute level and employs machine unlearning algorithms for the privacy preservation of model parameters. This user-interactive framework allows for adjustments in privacy protection intensity based on user feedback on generated images, striking a balance between maximal privacy safeguarding and maintaining model performance. Within this framework, we instantiate two modules: a differential privacy diffusion model for protecting attribute information in images and a feature unlearning algorithm for efficient updates of the trained model on the revised image dataset. Our approach demonstrated superiority over existing methods on facial datasets across various attribute classifications.
Authors: Yilei Qian, Kanglei Geng, Kailong Chen, Shaoxu Cheng, Linfeng Xu, Hongliang Li, Fanman Meng, Qingbo Wu
Abstract: The application of activity recognition in the "AI + Education" field is gaining increasing attention. However, current work mainly focuses on the recognition of activities in manually captured videos and a limited number of activity types, with little attention given to recognizing activities in surveillance images from real classrooms. In real classroom settings, normal teaching activities such as reading, account for a large proportion of samples, while rare non-teaching activities such as eating, continue to appear. This requires a model that can learn non-teaching activities from few samples without forgetting the normal teaching activities, which necessitates fewshot continual learning (FSCL) capability. To address this gap, we constructed a continual learning dataset focused on classroom surveillance image activity recognition called ARIC (Activity Recognition in Classroom). The dataset has advantages such as multiple perspectives, a wide variety of activities, and real-world scenarios, but it also presents challenges like similar activities and imbalanced sample distribution. To overcome these challenges, we designed a few-shot continual learning method that combines supervised contrastive learning (SCL) and an adaptive covariance classifier (ACC). During the base phase, we proposed a SCL approach based on feature augmentation to enhance the model's generalization ability. In the incremental phase, we employed an ACC to more accurately describe the distribution of new classes. Experimental results demonstrate that our method outperforms other existing methods on the ARIC dataset.
Authors: Friedhelm Hamann, Hanxiong Li, Paul Mieske, Lars Lewejohann, Guillermo Gallego
Abstract: Enabled by large annotated datasets, tracking and segmentation of objects in videos has made remarkable progress in recent years. Despite these advancements, algorithms still struggle under degraded conditions and during fast movements. Event cameras are novel sensors with high temporal resolution and high dynamic range that offer promising advantages to address these challenges. However, annotated data for developing learning-based mask-level tracking algorithms with events is not available. To this end, we introduce: ($i$) a new task termed \emph{space-time instance segmentation}, similar to video instance segmentation, whose goal is to segment instances throughout the entire duration of the sensor input (here, the input are quasi-continuous events and optionally aligned frames); and ($ii$) \emph{\dname}, a dataset for the new task, containing aligned grayscale frames and events. It includes annotated ground-truth labels (pixel-level instance segmentation masks) of a group of up to seven freely moving and interacting mice. We also provide two reference methods, which show that leveraging event data can consistently improve tracking performance, especially when used in combination with conventional cameras. The results highlight the potential of event-aided tracking in difficult scenarios. We hope our dataset opens the field of event-based video instance segmentation and enables the development of robust tracking algorithms for challenging conditions.\url{https://github.com/tub-rip/MouseSIS}
Authors: Jingcheng Ke, Dele Wang, Jun-Cheng Chen, I-Hong Jhuo, Chia-Wen Lin, Yen-Yu Lin
Abstract: One common belief is that with complex models and pre-training on large-scale datasets, transformer-based methods for referring expression comprehension (REC) perform much better than existing graph-based methods. We observe that since most graph-based methods adopt an off-the-shelf detector to locate candidate objects (i.e., regions detected by the object detector), they face two challenges that result in subpar performance: (1) the presence of significant noise caused by numerous irrelevant objects during reasoning, and (2) inaccurate localization outcomes attributed to the provided detector. To address these issues, we introduce a plug-and-adapt module guided by sub-expressions, called dynamic gate constraint (DGC), which can adaptively disable irrelevant proposals and their connections in graphs during reasoning. We further introduce an expression-guided regression strategy (EGR) to refine location prediction. Extensive experimental results on the RefCOCO, RefCOCO+, RefCOCOg, Flickr30K, RefClef, and Ref-reasoning datasets demonstrate the effectiveness of the DGC module and the EGR strategy in consistently boosting the performances of various graph-based REC methods. Without any pretaining, the proposed graph-based method achieves better performance than the state-of-the-art (SOTA) transformer-based methods.
Authors: Aoxiang Ning, Minglong Xue, Jinhong He, Chengyun Song
Abstract: Existing low-light image enhancement methods are difficult to fit the complex nonlinear relationship between normal and low-light images due to uneven illumination and noise effects. The recently proposed Kolmogorov-Arnold networks (KANs) feature spline-based convolutional layers and learnable activation functions, which can effectively capture nonlinear dependencies. In this paper, we design a KAN-Block based on KANs and innovatively apply it to low-light image enhancement. This method effectively alleviates the limitations of current methods constrained by linear network structures and lack of interpretability, further demonstrating the potential of KANs in low-level vision tasks. Given the poor perception of current low-light image enhancement methods and the stochastic nature of the inverse diffusion process, we further introduce frequency-domain perception for visually oriented enhancement. Extensive experiments demonstrate the competitive performance of our method on benchmark datasets. The code will be available at: https://github.com/AXNing/KSID}{https://github.com/AXNing/KSID.
URLs: https://github.com/AXNing/KSID, https://github.com/AXNing/KSID.
Authors: Yihao Zhao, Enhao Zhong, Cuiyun Yuan, Yang Li, Man Zhao, Chunxia Li, Jun Hu, Chenbin Liu
Abstract: We propose TG-LMM (Text-Guided Large Multi-Modal Model), a novel approach that leverages textual descriptions of organs to enhance segmentation accuracy in medical images. Existing medical image segmentation methods face several challenges: current medical automatic segmentation models do not effectively utilize prior knowledge, such as descriptions of organ locations; previous text-visual models focus on identifying the target rather than improving the segmentation accuracy; prior models attempt to use prior knowledge to enhance accuracy but do not incorporate pre-trained models. To address these issues, TG-LMM integrates prior knowledge, specifically expert descriptions of the spatial locations of organs, into the segmentation process. Our model utilizes pre-trained image and text encoders to reduce the number of training parameters and accelerate the training process. Additionally, we designed a comprehensive image-text information fusion structure to ensure thorough integration of the two modalities of data. We evaluated TG-LMM on three authoritative medical image datasets, encompassing the segmentation of various parts of the human body. Our method demonstrated superior performance compared to existing approaches, such as MedSAM, SAM and nnUnet.
Authors: Anwen Hu, Haiyang Xu, Liang Zhang, Jiabo Ye, Ming Yan, Ji Zhang, Qin Jin, Fei Huang, Jingren Zhou
Abstract: Multimodel Large Language Models(MLLMs) have achieved promising OCR-free Document Understanding performance by increasing the supported resolution of document images. However, this comes at the cost of generating thousands of visual tokens for a single document image, leading to excessive GPU memory and slower inference times, particularly in multi-page document comprehension. In this work, to address these challenges, we propose a High-resolution DocCompressor module to compress each high-resolution document image into 324 tokens, guided by low-resolution global visual features. With this compression module, to strengthen multi-page document comprehension ability and balance both token efficiency and question-answering performance, we develop the DocOwl2 under a three-stage training framework: Single-image Pretraining, Multi-image Continue-pretraining, and Multi-task Finetuning. DocOwl2 sets a new state-of-the-art across multi-page document understanding benchmarks and reduces first token latency by more than 50%, demonstrating advanced capabilities in multi-page questioning answering, explanation with evidence pages, and cross-page structure understanding. Additionally, compared to single-image MLLMs trained on similar data, our DocOwl2 achieves comparable single-page understanding performance with less than 20% of the visual tokens. Our codes, models, and data are publicly available at https://github.com/X-PLUG/mPLUG-DocOwl/tree/main/DocOwl2.
URLs: https://github.com/X-PLUG/mPLUG-DocOwl/tree/main/DocOwl2.
Authors: Hemanth Saratchandran, Thomas X. Wang, Simon Lucey
Abstract: In this article, we introduce a novel normalization technique for neural network weight matrices, which we term weight conditioning. This approach aims to narrow the gap between the smallest and largest singular values of the weight matrices, resulting in better-conditioned matrices. The inspiration for this technique partially derives from numerical linear algebra, where well-conditioned matrices are known to facilitate stronger convergence results for iterative solvers. We provide a theoretical foundation demonstrating that our normalization technique smoothens the loss landscape, thereby enhancing convergence of stochastic gradient descent algorithms. Empirically, we validate our normalization across various neural network architectures, including Convolutional Neural Networks (CNNs), Vision Transformers (ViT), Neural Radiance Fields (NeRF), and 3D shape modeling. Our findings indicate that our normalization method is not only competitive but also outperforms existing weight normalization techniques from the literature.
Authors: Lulin Li, Ben Chen, Xuechao Zou, Junliang Xing, Pin Tao
Abstract: Owing to the diverse geographical environments, intricate landscapes, and high-density settlements, the automatic identification of urban village boundaries using remote sensing images is a highly challenging task. This paper proposes a novel and efficient neural network model called UV-Mamba for accurate boundary detection in high-resolution remote sensing images. UV-Mamba mitigates the memory loss problem in long sequence modeling, which arises in state space model (SSM) with increasing image size, by incorporating deformable convolutions (DCN). Its architecture utilizes an encoder-decoder framework, includes an encoder with four deformable state space augmentation (DSSA) blocks for efficient multi-level semantic extraction and a decoder to integrate the extracted semantic information. We conducted experiments on the Beijing and Xi'an datasets, and the results show that UV-Mamba achieves state-of-the-art performance. Specifically, our model achieves 73.3% and 78.1% IoU on the Beijing and Xi'an datasets, respectively, representing improvements of 1.2% and 3.4% IoU over the previous best model, while also being 6x faster in inference speed and 40x smaller in parameter count. Source code and pre-trained models are available in the supplementary material.
Authors: Ibtissam Saadi, Douglas W. Cunningham, Taleb-ahmed Abdelmalik, Abdenour Hadid, Yassin El Hillali
Abstract: Existing methods for driver facial expression recognition (DFER) are often computationally intensive, rendering them unsuitable for real-time applications. In this work, we introduce a novel transfer learning-based dual architecture, named ShuffViT-DFER, which elegantly combines computational efficiency and accuracy. This is achieved by harnessing the strengths of two lightweight and efficient models using convolutional neural network (CNN) and vision transformers (ViT). We efficiently fuse the extracted features to enhance the performance of the model in accurately recognizing the facial expressions of the driver. Our experimental results on two benchmarking and public datasets, KMU-FED and KDEF, highlight the validity of our proposed method for real-time application with superior performance when compared to state-of-the-art methods.
Authors: Felix Sattler, Borja Carrillo Perez, Maurice Stephan, Sarah Barnes
Abstract: We introduce a novel method for updating 3D geospatial models, specifically targeting occlusion removal in large-scale maritime environments. Traditional 3D reconstruction techniques often face problems with dynamic objects, like cars or vessels, that obscure the true environment, leading to inaccurate models or requiring extensive manual editing. Our approach leverages deep learning techniques, including instance segmentation and generative inpainting, to directly modify both the texture and geometry of 3D meshes without the need for costly reprocessing. By selectively targeting occluding objects and preserving static elements, the method enhances both geometric and visual accuracy. This approach not only preserves structural and textural details of map data but also maintains compatibility with current geospatial standards, ensuring robust performance across diverse datasets. The results demonstrate significant improvements in 3D model fidelity, making this method highly applicable for maritime situational awareness and the dynamic display of auxiliary information.
Authors: Pei Wang, Xiaotong Luo, Yuan Xie, Yanyun Qu
Abstract: Multi-weather image restoration has witnessed incredible progress, while the increasing model capacity and expensive data acquisition impair its applications in memory-limited devices. Data-free distillation provides an alternative for allowing to learn a lightweight student model from a pre-trained teacher model without relying on the original training data. The existing data-free learning methods mainly optimize the models with the pseudo data generated by GANs or the real data collected from the Internet. However, they inevitably suffer from the problems of unstable training or domain shifts with the original data. In this paper, we propose a novel Data-free Distillation with Degradation-prompt Diffusion framework for multi-weather Image Restoration (D4IR). It replaces GANs with pre-trained diffusion models to avoid model collapse and incorporates a degradation-aware prompt adapter to facilitate content-driven conditional diffusion for generating domain-related images. Specifically, a contrast-based degradation prompt adapter is firstly designed to capture degradation-aware prompts from web-collected degraded images. Then, the collected unpaired clean images are perturbed to latent features of stable diffusion, and conditioned with the degradation-aware prompts to synthesize new domain-related degraded images for knowledge distillation. Experiments illustrate that our proposal achieves comparable performance to the model distilled with original training data, and is even superior to other mainstream unsupervised methods.
Authors: Hanyang Yu, Xiaoxiao Long, Ping Tan
Abstract: We aim to address sparse-view reconstruction of a 3D scene by leveraging priors from large-scale vision models. While recent advancements such as 3D Gaussian Splatting (3DGS) have demonstrated remarkable successes in 3D reconstruction, these methods typically necessitate hundreds of input images that densely capture the underlying scene, making them time-consuming and impractical for real-world applications. However, sparse-view reconstruction is inherently ill-posed and under-constrained, often resulting in inferior and incomplete outcomes. This is due to issues such as failed initialization, overfitting on input images, and a lack of details. To mitigate these challenges, we introduce LM-Gaussian, a method capable of generating high-quality reconstructions from a limited number of images. Specifically, we propose a robust initialization module that leverages stereo priors to aid in the recovery of camera poses and the reliable point clouds. Additionally, a diffusion-based refinement is iteratively applied to incorporate image diffusion priors into the Gaussian optimization process to preserve intricate scene details. Finally, we utilize video diffusion priors to further enhance the rendered images for realistic visual effects. Overall, our approach significantly reduces the data acquisition requirements compared to previous 3DGS methods. We validate the effectiveness of our framework through experiments on various public datasets, demonstrating its potential for high-quality 360-degree scene reconstruction. Visual results are on our website.
Authors: Akshay Jain, Shiv Ram Dubey, Satish Kumar Singh, KC Santosh, Bidyut Baran Chaudhuri
Abstract: Convolutional Neural Networks (CNNs) have made remarkable strides; however, they remain susceptible to vulnerabilities, particularly in the face of minor image perturbations that humans can easily recognize. This weakness, often termed as 'attacks', underscores the limited robustness of CNNs and the need for research into fortifying their resistance against such manipulations. This study introduces a novel Non-Uniform Illumination (NUI) attack technique, where images are subtly altered using varying NUI masks. Extensive experiments are conducted on widely-accepted datasets including CIFAR10, TinyImageNet, and CalTech256, focusing on image classification with 12 different NUI attack models. The resilience of VGG, ResNet, MobilenetV3-small and InceptionV3 models against NUI attacks are evaluated. Our results show a substantial decline in the CNN models' classification accuracy when subjected to NUI attacks, indicating their vulnerability under non-uniform illumination. To mitigate this, a defense strategy is proposed, including NUI-attacked images, generated through the new NUI transformation, into the training set. The results demonstrate a significant enhancement in CNN model performance when confronted with perturbed images affected by NUI attacks. This strategy seeks to bolster CNN models' resilience against NUI attacks.
Authors: Moritz Nottebaum, Matteo Dunnhofer, Christian Micheloni
Abstract: Research in efficient vision backbones is evolving into models that are a mixture of convolutions and transformer blocks. A smart combination of both, architecture-wise and component-wise is mandatory to excel in the speedaccuracy trade-off. Most publications focus on maximizing accuracy and utilize MACs (multiply accumulate operations) as an efficiency metric. The latter however often do not measure accurately how fast a model actually is due to factors like memory access cost and degree of parallelism. We analyzed common modules and architectural design choices for backbones not in terms of MACs, but rather in actual throughput and latency, as the combination of the latter two is a better representation of the efficiency of models in real applications. We applied the conclusions taken from that analysis to create a recipe for increasing hardware-efficiency in macro design. Additionally we introduce a simple slimmed-down version of MultiHead Self-Attention, that aligns with our analysis. We combine both macro and micro design to create a new family of hardware-efficient backbone networks called LowFormer. LowFormer achieves a remarkable speedup in terms of throughput and latency, while achieving similar or better accuracy than current state-of-the-art efficient backbones. In order to prove the generalizability of our hardware-efficient design, we evaluate our method on GPU, mobile GPU and ARM CPU. We further show that the downstream tasks object detection and semantic segmentation profit from our hardware-efficient architecture. Code and models are available at https://github.com/ altair199797/LowFormer.
URLs: https://github.com/
Authors: Prerak Mody, Nicolas F. Chaves-de-Plaza, Chinmay Rao, Eleftheria Astrenidou, Mischa de Ridder, Nienke Hoekstra, Klaus Hildebrandt, Marius Staring
Abstract: Increased usage of automated tools like deep learning in medical image segmentation has alleviated the bottleneck of manual contouring. This has shifted manual labour to quality assessment (QA) of automated contours which involves detecting errors and correcting them. A potential solution to semi-automated QA is to use deep Bayesian uncertainty to recommend potentially erroneous regions, thus reducing time spent on error detection. Previous work has investigated the correspondence between uncertainty and error, however, no work has been done on improving the "utility" of Bayesian uncertainty maps such that it is only present in inaccurate regions and not in the accurate ones. Our work trains the FlipOut model with the Accuracy-vs-Uncertainty (AvU) loss which promotes uncertainty to be present only in inaccurate regions. We apply this method on datasets of two radiotherapy body sites, c.f. head-and-neck CT and prostate MR scans. Uncertainty heatmaps (i.e. predictive entropy) are evaluated against voxel inaccuracies using Receiver Operating Characteristic (ROC) and Precision-Recall (PR) curves. Numerical results show that when compared to the Bayesian baseline the proposed method successfully suppresses uncertainty for accurate voxels, with similar presence of uncertainty for inaccurate voxels. Code to reproduce experiments is available at https://github.com/prerakmody/bayesuncertainty-error-correspondence
URLs: https://github.com/prerakmody/bayesuncertainty-error-correspondence
Authors: Xiujian Liang, Gaozhi Liu, Yichao Si, Xiaoxiao Hu, Zhenxing Qian, Xinpeng Zhang
Abstract: Digital watermarking has demonstrated its effectiveness in protecting multimedia content. However, existing watermarking are predominantly tailored for specific media types, rendering them less effective for the protection of content displayed on computer screens, which is often multimodal and dynamic. Visual Screen Content (VSC), is particularly susceptible to theft and leakage via screenshots, a vulnerability that current watermarking methods fail to adequately address.To tackle these challenges, we propose ScreenMark, a robust and practical watermarking method designed specifically for arbitrary VSC protection. ScreenMark utilizes a three-stage progressive watermarking framework. Initially, inspired by diffusion principles, we initialize the mutual transformation between regular watermark information and irregular watermark patterns. Subsequently, these patterns are integrated with screen content using a pre-multiplication alpha blending technique, supported by a pre-trained screen decoder for accurate watermark retrieval. The progressively complex distorter enhances the robustness of the watermark in real-world screenshot scenarios. Finally, the model undergoes fine-tuning guided by a joint-level distorter to ensure optimal performance.To validate the effectiveness of ScreenMark, we compiled a dataset comprising 100,000 screenshots from various devices and resolutions. Extensive experiments across different datasets confirm the method's superior robustness, imperceptibility, and practical applicability.
Authors: Rizhao Cai, Cecelia Soh, Zitong Yu, Haoliang Li, Wenhan Yang, Alex Kot
Abstract: Face Anti-Spoofing (FAS) research is challenged by the cross-domain problem, where there is a domain gap between the training and testing data. While recent FAS works are mainly model-centric, focusing on developing domain generalization algorithms for improving cross-domain performance, data-centric research for face anti-spoofing, improving generalization from data quality and quantity, is largely ignored. Therefore, our work starts with data-centric FAS by conducting a comprehensive investigation from the data perspective for improving cross-domain generalization of FAS models. More specifically, at first, based on physical procedures of capturing and recapturing, we propose task-specific FAS data augmentation (FAS-Aug), which increases data diversity by synthesizing data of artifacts, such as printing noise, color distortion, moir\'e pattern, \textit{etc}. Our experiments show that using our FAS augmentation can surpass traditional image augmentation in training FAS models to achieve better cross-domain performance. Nevertheless, we observe that models may rely on the augmented artifacts, which are not environment-invariant, and using FAS-Aug may have a negative effect. As such, we propose Spoofing Attack Risk Equalization (SARE) to prevent models from relying on certain types of artifacts and improve the generalization performance. Last but not least, our proposed FAS-Aug and SARE with recent Vision Transformer backbones can achieve state-of-the-art performance on the FAS cross-domain generalization protocols. The implementation is available at https://github.com/RizhaoCai/FAS_Aug.
Authors: Chamuditha Jayanaga Galappaththige, Zachary Izzo, Xilin He, Honglu Zhou, Muhammad Haris Khan
Abstract: Unarguably, deep learning models capable of generalizing to unseen domain data while leveraging a few labels are of great practical significance due to low developmental costs. In search of this endeavor, we study the challenging problem of semi-supervised domain generalization (SSDG), where the goal is to learn a domain-generalizable model while using only a small fraction of labeled data and a relatively large fraction of unlabeled data. Domain generalization (DG) methods show subpar performance under the SSDG setting, whereas semi-supervised learning (SSL) methods demonstrate relatively better performance, however, they are considerably poor compared to the fully-supervised DG methods. Towards handling this new, but challenging problem of SSDG, we propose a novel method that can facilitate the generation of accurate pseudo-labels under various domain shifts. This is accomplished by retaining the domain-level specialism in the classifier during training corresponding to each source domain. Specifically, we first create domain-level information vectors on the fly which are then utilized to learn a domain-aware mask for modulating the classifier's weights. We provide a mathematical interpretation for the effect of this modulation procedure on both pseudo-labeling and model training. Our method is plug-and-play and can be readily applied to different SSL baselines for SSDG. Extensive experiments on six challenging datasets in two different SSDG settings show that our method provides visible gains over the various strong SSL-based SSDG baselines.
Authors: Deyin Liu, Lin Yuanbo Wu, Xianghua Xie
Abstract: Due to lack of fully publicly available text-to-video models, current video editing methods tend to build on pre-trained text-to-image generation models, however, they still face grand challenges in dealing with the local editing of video with temporal information. First, although existing methods attempt to focus on local area editing by a pre-defined mask, the preservation of the outside-area background is non-ideal due to the spatially entire generation of each frame. In addition, specially providing a mask by user is an additional costly undertaking, so an autonomous masking strategy integrated into the editing process is desirable. Last but not least, image-level pretrained model hasn't learned temporal information across frames of a video which is vital for expressing the motion and dynamics. In this paper, we propose to adapt a image-level blended latent diffusion model to perform local video editing tasks. Specifically, we leverage DDIM inversion to acquire the latents as background latents instead of the randomly noised ones to better preserve the background information of the input video. We further introduce an autonomous mask manufacture mechanism derived from cross-attention maps in diffusion steps. Finally, we enhance the temporal consistency across video frames by transforming the self-attention blocks of U-Net into temporal-spatial blocks. Through extensive experiments, our proposed approach demonstrates effectiveness in different real-world video editing tasks.
Authors: Jeongsoo Kim, Jongho Nang, Junsuk Choe
Abstract: Recent Vision Transformer (ViT)-based methods for Image Super-Resolution have demonstrated impressive performance. However, they suffer from significant complexity, resulting in high inference times and memory usage. Additionally, ViT models using Window Self-Attention (WSA) face challenges in processing regions outside their windows. To address these issues, we propose the Low-to-high Multi-Level Transformer (LMLT), which employs attention with varying feature sizes for each head. LMLT divides image features along the channel dimension, gradually reduces spatial size for lower heads, and applies self-attention to each head. This approach effectively captures both local and global information. By integrating the results from lower heads into higher heads, LMLT overcomes the window boundary issues in self-attention. Extensive experiments show that our model significantly reduces inference time and GPU memory usage while maintaining or even surpassing the performance of state-of-the-art ViT-based Image Super-Resolution methods. Our codes are availiable at https://github.com/jwgdmkj/LMLT.
Authors: Ombretta Strafforello, Derya Soydaner, Michiel Willems, Anne-Sofie Maerten, Stefanie De Winter
Abstract: The emergence of large Vision-Language Models (VLMs) has recently established new baselines in image classification across multiple domains. However, the performance of VLMs in the specific task of artwork classification, particularly art style classification of paintings - a domain traditionally mastered by art historians - has not been explored yet. Artworks pose a unique challenge compared to natural images due to their inherently complex and diverse structures, characterized by variable compositions and styles. Art historians have long studied the unique aspects of artworks, with style prediction being a crucial component of their discipline. This paper investigates whether large VLMs, which integrate visual and textual data, can effectively predict the art historical attributes of paintings. We conduct an in-depth analysis of four VLMs, namely CLIP, LLaVA, OpenFlamingo, and GPT-4o, focusing on zero-shot classification of art style, author and time period using two public benchmarks of artworks. Additionally, we present ArTest, a well-curated test set of artworks, including pivotal paintings studied by art historians.
Authors: Xi Chen, Haosen Yang, Sheng Jin, Xiatian Zhu, Hongxun Yao
Abstract: Open-vocabulary segmentation poses significant challenges, as it requires segmenting and recognizing objects across an open set of categories in unconstrained environments. Building on the success of powerful vision-language (ViL) foundation models, such as CLIP, recent efforts sought to harness their zero-short capabilities to recognize unseen categories. Despite notable performance improvements, these models still encounter the critical issue of generating precise mask proposals for unseen categories and scenarios, resulting in inferior segmentation performance eventually. To address this challenge, we introduce a novel approach, FrozenSeg, designed to integrate spatial knowledge from a localization foundation model (e.g., SAM) and semantic knowledge extracted from a ViL model (e.g., CLIP), in a synergistic framework. Taking the ViL model's visual encoder as the feature backbone, we inject the space-aware feature into the learnable queries and CLIP features within the transformer decoder. In addition, we devise a mask proposal ensemble strategy for further improving the recall rate and mask quality. To fully exploit pre-trained knowledge while minimizing training overhead, we freeze both foundation models, focusing optimization efforts solely on a lightweight transformer decoder for mask proposal generation-the performance bottleneck. Extensive experiments demonstrate that FrozenSeg advances state-of-the-art results across various segmentation benchmarks, trained exclusively on COCO panoptic data, and tested in a zero-shot manner. Code is available at https://github.com/chenxi52/FrozenSeg.
Authors: Sebastian Pulgar, Domingo Mery
Abstract: In recent years, facial recognition (FR) models have become the most widely used biometric tool, achieving impressive results on numerous datasets. However, inherent hardware challenges or shooting distances often result in low-resolution images, which significantly impact the performance of FR models. To address this issue, several solutions have been proposed, including super-resolution (SR) models that generate highly realistic faces. Despite these efforts, significant improvements in FR algorithms have not been achieved. We propose a novel SR model FTLGAN, which focuses on generating high-resolution images that preserve individual identities rather than merely improving image quality, thereby maximizing the performance of FR models. The results are compelling, demonstrating a mean value of d' 21% above the best current state-of-the-art models, specifically having a value of d' = 1.099 and AUC = 0.78 for 14x14 pixels, d' = 2.112 and AUC = 0.92 for 28x28 pixels, and d' = 3.049 and AUC = 0.98 for 56x56 pixels. The contributions of this study are significant in several key areas. Firstly, a notable improvement in facial recognition performance has been achieved in low-resolution images, specifically at resolutions of 14x14, 28x28, and 56x56 pixels. Secondly, the enhancements demonstrated by FTLGAN show a consistent response across all resolutions, delivering outstanding performance uniformly, unlike other comparative models. Thirdly, an innovative approach has been implemented using triplet loss logic, enabling the training of the super-resolution model solely with real images, contrasting with current models, and expanding potential real-world applications. Lastly, this study introduces a novel model that specifically addresses the challenge of improving classification performance in facial recognition systems by integrating facial recognition quality as a loss during model training.
Authors: Fabian Diet, Moussa Kassem Sbeyti, Michelle Karg
Abstract: Natural distribution shift causes a deterioration in the perception performance of convolutional neural networks (CNNs). This comprehensive analysis for real-world traffic data addresses: 1) investigating the effect of natural distribution shift and weather augmentations on both detection quality and confidence estimation, 2) evaluating model performance for both classification and object localization, and 3) benchmarking two common uncertainty quantification methods - Ensembles and different variants of Monte-Carlo (MC) Dropout - under natural and close-to-natural distribution shift. For this purpose, a novel dataset has been curated from publicly available autonomous driving datasets. The in-distribution (ID) data is based on cutouts of a single object, for which both class and bounding box annotations are available. The six distribution-shift datasets cover adverse weather scenarios, simulated rain and fog, corner cases, and out-of-distribution data. A granular analysis of CNNs under distribution shift allows to quantize the impact of different types of shifts on both, task performance and confidence estimation: ConvNeXt-Tiny is more robust than EfficientNet-B0; heavy rain degrades classification stronger than localization, contrary to heavy fog; integrating MC-Dropout into selected layers only has the potential to enhance task performance and confidence estimation, whereby the identification of these layers depends on the type of distribution shift and the considered task.
Authors: Qianlong Xiang, Miao Zhang, Yuzhang Shang, Jianlong Wu, Yan Yan, Liqiang Nie
Abstract: Diffusion models (DMs) have demonstrated exceptional generative capabilities across various areas, while they are hindered by slow inference speeds and high computational demands during deployment. The most common way to accelerate DMs involves reducing the number of denoising steps during generation, achieved through faster sampling solvers or knowledge distillation (KD). In contrast to prior approaches, we propose a novel method that transfers the capability of large pretrained DMs to faster architectures. Specifically, we employ KD in a distinct manner to compress DMs by distilling their generative ability into more rapid variants. Furthermore, considering that the source data is either unaccessible or too enormous to store for current generative models, we introduce a new paradigm for their distillation without source data, termed Data-Free Knowledge Distillation for Diffusion Models (DKDM). Generally, our established DKDM framework comprises two main components: 1) a DKDM objective that uses synthetic denoising data produced by pretrained DMs to optimize faster DMs without source data, and 2) a dynamic iterative distillation method that flexibly organizes the synthesis of denoising data, preventing it from slowing down the optimization process as the generation is slow. To our knowledge, this is the first attempt at using KD to distill DMs into any architecture in a data-free manner. Importantly, our DKDM is orthogonal to most existing acceleration methods, such as denoising step reduction, quantization and pruning. Experiments show that our DKDM is capable of deriving 2x faster DMs with performance remaining on par with the baseline. Notably, our DKDM enables pretrained DMs to function as "datasets" for training new DMs.
Authors: Rongzhen Zhao, Vivienne Wang, Juho Kannala, Joni Pajarinen
Abstract: Object-Centric Learning (OCL) represents dense image or video pixels as sparse object features. Representative methods utilize discrete representation composed of Variational Autoencoder (VAE) template features to suppress pixel-level information redundancy and guide object-level feature aggregation. The most recent advancement, Grouped Discrete Representation (GDR), further decomposes these template features into attributes. However, its naive channel grouping as decomposition may erroneously group channels belonging to different attributes together and discretize them as sub-optimal template attributes, which losses information and harms expressivity. We propose Organized GDR (OGDR) to organize channels belonging to the same attributes together for correct decomposition from features into attributes. In unsupervised segmentation experiments, OGDR is fully superior to GDR in augmentating classical transformer-based OCL methods; it even improves state-of-the-art diffusion-based ones. Codebook PCA and representation similarity analyses show that compared with GDR, our OGDR eliminates redundancy and preserves information better for guiding object representation learning. The source code is available in the supplementary material.
Authors: Richard Franklin, Jiawei Yao, Deyang Zhong, Qi Qian, Juhua Hu
Abstract: In many real-world applications, the frequency distribution of class labels for training data can exhibit a long-tailed distribution, which challenges traditional approaches of training deep neural networks that require heavy amounts of balanced data. Gathering and labeling data to balance out the class label distribution can be both costly and time-consuming. Many existing solutions that enable ensemble learning, re-balancing strategies, or fine-tuning applied to deep neural networks are limited by the inert problem of few class samples across a subset of classes. Recently, vision-language models like CLIP have been observed as effective solutions to zero-shot or few-shot learning by grasping a similarity between vision and language features for image and text pairs. Considering that large pre-trained vision-language models may contain valuable side textual information for minor classes, we propose to leverage text supervision to tackle the challenge of long-tailed learning. Concretely, we propose a novel text-guided mixup technique that takes advantage of the semantic relations between classes recognized by the pre-trained text encoder to help alleviate the long-tailed problem. Our empirical study on benchmark long-tailed tasks demonstrates the effectiveness of our proposal with a theoretical guarantee. Our code is available at https://github.com/rsamf/text-guided-mixup.
Authors: Bernardo Biesseck, Pedro Vidal, Luiz Coelho, Roger Granada, David Menotti|
Abstract: A robust face recognition model must be trained using datasets that include a large number of subjects and numerous samples per subject under varying conditions (such as pose, expression, age, noise, and occlusion). Due to ethical and privacy concerns, large-scale real face datasets have been discontinued, such as MS1MV3, and synthetic face generators have been proposed, utilizing GANs and Diffusion Models, such as SYNFace, SFace, DigiFace-1M, IDiff-Face, DCFace, and GANDiffFace, aiming to supply this demand. Some of these methods can produce high-fidelity realistic faces, but with low intra-class variance, while others generate high-variance faces with low identity consistency. In this paper, we propose a Triple Condition Diffusion Model (TCDiff) to improve face style transfer from real to synthetic faces through 2D and 3D facial constraints, enhancing face identity consistency while keeping the necessary high intra-class variance. Face recognition experiments using 1k, 2k, and 5k classes of our new dataset for training outperform state-of-the-art synthetic datasets in real face benchmarks such as LFW, CFP-FP, AgeDB, and BUPT. Our source code is available at: https://github.com/BOVIFOCR/tcdiff.
Authors: Lingyu Xiong, Xize Cheng, Jintao Tan, Xianjia Wu, Xiandong Li, Lei Zhu, Fei Ma, Minglei Li, Huang Xu, Zhihu Hu
Abstract: Audio-driven talking face generation aims to synthesize video with lip movements synchronized to input audio. However, current generative techniques face challenges in preserving intricate regional textures (skin, teeth). To address the aforementioned challenges, we propose a novel framework called SegTalker to decouple lip movements and image textures by introducing segmentation as intermediate representation. Specifically, given the mask of image employed by a parsing network, we first leverage the speech to drive the mask and generate talking segmentation. Then we disentangle semantic regions of image into style codes using a mask-guided encoder. Ultimately, we inject the previously generated talking segmentation and style codes into a mask-guided StyleGAN to synthesize video frame. In this way, most of textures are fully preserved. Moreover, our approach can inherently achieve background separation and facilitate mask-guided facial local editing. In particular, by editing the mask and swapping the region textures from a given reference image (e.g. hair, lip, eyebrows), our approach enables facial editing seamlessly when generating talking face video. Experiments demonstrate that our proposed approach can effectively preserve texture details and generate temporally consistent video while remaining competitive in lip synchronization. Quantitative and qualitative results on the HDTF and MEAD datasets illustrate the superior performance of our method over existing methods.
Authors: Rui Peng, Shihe Shen, Kaiqiang Xiong, Huachen Gao, Jianbo Jiao, Xiaodong Gu, Ronggang Wang
Abstract: Reconstructing the high-fidelity surface from multi-view images, especially sparse images, is a critical and practical task that has attracted widespread attention in recent years. However, existing methods are impeded by the memory constraint or the requirement of ground-truth depths and cannot recover satisfactory geometric details. To this end, we propose SuRF, a new Surface-centric framework that incorporates a new Region sparsification based on a matching Field, achieving good trade-offs between performance, efficiency and scalability. To our knowledge, this is the first unsupervised method achieving end-to-end sparsification powered by the introduced matching field, which leverages the weight distribution to efficiently locate the boundary regions containing surface. Instead of predicting an SDF value for each voxel, we present a new region sparsification approach to sparse the volume by judging whether the voxel is inside the surface region. In this way, our model can exploit higher frequency features around the surface with less memory and computational consumption. Extensive experiments on multiple benchmarks containing complex large-scale scenes show that our reconstructions exhibit high-quality details and achieve new state-of-the-art performance, i.e., 46% improvements with 80% less memory consumption. Code is available at https://github.com/prstrive/SuRF.
Authors: Bin Wang, Fan Wu, Linke Ouyang, Zhuangcheng Gu, Rui Zhang, Renqiu Xia, Bo Zhang, Conghui He
Abstract: Formula recognition presents significant challenges due to the complicated structure and varied notation of mathematical expressions. Despite continuous advancements in formula recognition models, the evaluation metrics employed by these models, such as BLEU and Edit Distance, still exhibit notable limitations. They overlook the fact that the same formula has diverse representations and is highly sensitive to the distribution of training data, thereby causing the unfairness in formula recognition evaluation. To this end, we propose a Character Detection Matching (CDM) metric, ensuring the evaluation objectivity by designing a image-level rather than LaTex-level metric score. Specifically, CDM renders both the model-predicted LaTeX and the ground-truth LaTeX formulas into image-formatted formulas, then employs visual feature extraction and localization techniques for precise character-level matching, incorporating spatial position information. Such a spatially-aware and character-matching method offers a more accurate and equitable evaluation compared with previous BLEU and Edit Distance metrics that rely solely on text-based character matching. Experimentally, we evaluated various formula recognition models using CDM, BLEU, and ExpRate metrics. Their results demonstrate that the CDM aligns more closely with human evaluation standards and provides a fairer comparison across different models by eliminating discrepancies caused by diverse formula representations.
Authors: Benzhi Wang, Jingkai Zhou, Jingqi Bai, Yang Yang, Weihua Chen, Fan Wang, Zhen Lei
Abstract: In recent years, diffusion models have revolutionized visual generation, outperforming traditional frameworks like Generative Adversarial Networks (GANs). However, generating images of humans with realistic semantic parts, such as hands and faces, remains a significant challenge due to their intricate structural complexity. To address this issue, we propose a novel post-processing solution named RealisHuman. The RealisHuman framework operates in two stages. First, it generates realistic human parts, such as hands or faces, using the original malformed parts as references, ensuring consistent details with the original image. Second, it seamlessly integrates the rectified human parts back into their corresponding positions by repainting the surrounding areas to ensure smooth and realistic blending. The RealisHuman framework significantly enhances the realism of human generation, as demonstrated by notable improvements in both qualitative and quantitative metrics. Code is available at https://github.com/Wangbenzhi/RealisHuman.
Authors: Slava Elizarov, Ciara Rowles, Simon Donn\'e
Abstract: Generating high-quality 3D objects from textual descriptions remains a challenging problem due to computational cost, the scarcity of 3D data, and complex 3D representations. We introduce Geometry Image Diffusion (GIMDiffusion), a novel Text-to-3D model that utilizes geometry images to efficiently represent 3D shapes using 2D images, thereby avoiding the need for complex 3D-aware architectures. By integrating a Collaborative Control mechanism, we exploit the rich 2D priors of existing Text-to-Image models such as Stable Diffusion. This enables strong generalization even with limited 3D training data (allowing us to use only high-quality training data) as well as retaining compatibility with guidance techniques such as IPAdapter. In short, GIMDiffusion enables the generation of 3D assets at speeds comparable to current Text-to-Image models. The generated objects consist of semantically meaningful, separate parts and include internal structures, enhancing both usability and versatility.
Authors: Shuya Yang, Shaozhe Hao, Yukang Cao, Kwan-Yee K. Wong
Abstract: Subject-driven text-to-image generation has witnessed remarkable advancements in its ability to learn and capture characteristics of a subject using only a limited number of images. However, existing methods commonly rely on high-quality images for training and may struggle to generate reasonable images when the input images are blemished by artifacts. This is primarily attributed to the inadequate capability of current techniques in distinguishing subject-related features from disruptive artifacts. In this paper, we introduce ArtiFade to tackle this issue and successfully generate high-quality artifact-free images from blemished datasets. Specifically, ArtiFade exploits fine-tuning of a pre-trained text-to-image model, aiming to remove artifacts. The elimination of artifacts is achieved by utilizing a specialized dataset that encompasses both unblemished images and their corresponding blemished counterparts during fine-tuning. ArtiFade also ensures the preservation of the original generative capabilities inherent within the diffusion model, thereby enhancing the overall performance of subject-driven methods in generating high-quality and artifact-free images. We further devise evaluation benchmarks tailored for this task. Through extensive qualitative and quantitative experiments, we demonstrate the generalizability of ArtiFade in effective artifact removal under both in-distribution and out-of-distribution scenarios.
Authors: Marga Don, Stijn Pinson, Blanca Guillen Cebrian, Yuki M. Asano
Abstract: Foundation models (FMs) are a popular topic of research in AI. Their ability to generalize to new tasks and datasets without retraining or needing an abundance of data makes them an appealing candidate for applications on specialist datasets. In this work, we compare the performance of FMs to finetuned pre-trained supervised models in the task of semantic segmentation on an entirely new dataset. We see that finetuned models consistently outperform the FMs tested, even in cases were data is scarce. We release the code and dataset for this work on GitHub.
Authors: Wenliang Zhao, Haolin Wang, Jie Zhou, Jiwen Lu
Abstract: Diffusion probabilistic models (DPMs) have shown remarkable performance in visual synthesis but are computationally expensive due to the need for multiple evaluations during the sampling. Recent predictor-corrector diffusion samplers have significantly reduced the required number of function evaluations (NFE), but inherently suffer from a misalignment issue caused by the extra corrector step, especially with a large classifier-free guidance scale (CFG). In this paper, we introduce a new fast DPM sampler called DC-Solver, which leverages dynamic compensation (DC) to mitigate the misalignment of the predictor-corrector samplers. The dynamic compensation is controlled by compensation ratios that are adaptive to the sampling steps and can be optimized on only 10 datapoints by pushing the sampling trajectory toward a ground truth trajectory. We further propose a cascade polynomial regression (CPR) which can instantly predict the compensation ratios on unseen sampling configurations. Additionally, we find that the proposed dynamic compensation can also serve as a plug-and-play module to boost the performance of predictor-only samplers. Extensive experiments on both unconditional sampling and conditional sampling demonstrate that our DC-Solver can consistently improve the sampling quality over previous methods on different DPMs with a wide range of resolutions up to 1024$\times$1024. Notably, we achieve 10.38 FID (NFE=5) on unconditional FFHQ and 0.394 MSE (NFE=5, CFG=7.5) on Stable-Diffusion-2.1. Code is available at https://github.com/wl-zhao/DC-Solver
Authors: Yunze Man, Shuhong Zheng, Zhipeng Bao, Martial Hebert, Liang-Yan Gui, Yu-Xiong Wang
Abstract: Complex 3D scene understanding has gained increasing attention, with scene encoding strategies playing a crucial role in this success. However, the optimal scene encoding strategies for various scenarios remain unclear, particularly compared to their image-based counterparts. To address this issue, we present a comprehensive study that probes various visual encoding models for 3D scene understanding, identifying the strengths and limitations of each model across different scenarios. Our evaluation spans seven vision foundation encoders, including image-based, video-based, and 3D foundation models. We evaluate these models in four tasks: Vision-Language Scene Reasoning, Visual Grounding, Segmentation, and Registration, each focusing on different aspects of scene understanding. Our evaluations yield key findings: DINOv2 demonstrates superior performance, video models excel in object-level tasks, diffusion models benefit geometric tasks, and language-pretrained models show unexpected limitations in language-related tasks. These insights challenge some conventional understandings, provide novel perspectives on leveraging visual foundation models, and highlight the need for more flexible encoder selection in future vision-language and scene-understanding tasks.
Authors: Abby Stylianou, Michelle Brachman, Albatool Wazzan, Samuel Black, Richard Souvenir
Abstract: The data that underlies automated methods in computer vision and machine learning, such as image retrieval and fine-grained recognition, often comes from crowdsourcing. In contexts that rely on the intrinsic motivation of users, we seek to understand how the application design affects a user's willingness to contribute and the quantity and quality of the data they capture. In this project, we designed three versions of a camera-based mobile crowdsourcing application, which varied in the amount of labeling effort requested of the user and conducted a user study to evaluate the trade-off between the level of user-contributed information requested and the quantity and quality of labeled images collected. The results suggest that higher levels of user labeling do not lead to reduced contribution. Users collected and annotated the most images using the application version with the highest requested level of labeling with no decrease in user satisfaction. In preliminary experiments, the additional labeled data supported increased performance on an image retrieval task.
Authors: Amir Syahmi, Xiangrong Lu, Yinxuan Li, Haoxuan Yao, Hanjun Jiang, Ishita Acharya, Shiyi Wang, Yang Nan, Xiaodan Xing, Guang Yang
Abstract: Recent advancements in medical imaging and artificial intelligence (AI) have greatly enhanced diagnostic capabilities, but the development of effective deep learning (DL) models is still constrained by the lack of high-quality annotated datasets. The traditional manual annotation process by medical experts is time- and resource-intensive, limiting the scalability of these datasets. In this work, we introduce a robust and versatile framework that combines AI and crowdsourcing to improve both the quality and quantity of medical image datasets across different modalities. Our approach utilises a user-friendly online platform that enables a diverse group of crowd annotators to label medical images efficiently. By integrating the MedSAM segmentation AI with this platform, we accelerate the annotation process while maintaining expert-level quality through an algorithm that merges crowd-labelled images. Additionally, we employ pix2pixGAN, a generative AI model, to expand the training dataset with synthetic images that capture realistic morphological features. These methods are combined into a cohesive framework designed to produce an enhanced dataset, which can serve as a universal pre-processing pipeline to boost the training of any medical deep learning segmentation model. Our results demonstrate that this framework significantly improves model performance, especially when training data is limited.
Authors: Tahsin Reasat, Stephen Chenard, Akhil Rekulapelli, Nicholas Chadwick, Joanna Shechtel, Katherine van Schaik, David S. Smith, Joshua Lawrenz
Abstract: Accurate musculoskeletal soft tissue tumor segmentation is vital for assessing tumor size, location, diagnosis, and response to treatment, thereby influencing patient outcomes. However, segmentation of these tumors requires clinical expertise, and an automated segmentation model would save valuable time for both clinician and patient. Training an automatic model requires a large dataset of annotated images. In this work, we describe the collection of an MR imaging dataset of 199 musculoskeletal soft tissue tumors from 199 patients. We trained segmentation models on this dataset and then benchmarked them on a publicly available dataset. Our model achieved the state-of-the-art dice score of 0.79 out of the box without any fine tuning, which shows the diversity and utility of our curated dataset. We analyzed the model predictions and found that its performance suffered on fibrous and vascular tumors due to their diverse anatomical location, size, and intensity heterogeneity. The code and models are available in the following github repository, https://github.com/Reasat/mstt
Authors: Be\~nat Froemming-Aldanondo, Tatiana Rastoskueva, Michael Evans, Marcial Machado, Anna Vadella, Rickey Johnson, Luis Escamilla, Milan Jostes, Devson Butani, Ryan Kaddis, Chan-Jin Chung, Joshua Siegel
Abstract: Reliable lane-following algorithms are essential for safe and effective autonomous driving. This project was primarily focused on developing and evaluating different lane-following programs to find the most reliable algorithm for a Vehicle to Everything (V2X) project. The algorithms were first tested on a simulator and then with real vehicles equipped with a drive-by-wire system using ROS (Robot Operating System). Their performance was assessed through reliability, comfort, speed, and adaptability metrics. The results show that the two most reliable approaches detect both lane lines and use unsupervised learning to separate them. These approaches proved to be robust in various driving scenarios, making them suitable candidates for integration into the V2X project.
Authors: Qiwen Zhu, Yanjie Wang, Shilv Cai, Liqun Chen, Jiahuan Zhou, Luxin Yan, Sheng Zhong, Xu Zou
Abstract: Training Single-Image Super-Resolution (SISR) models using pixel-based regression losses can achieve high distortion metrics scores (e.g., PSNR and SSIM), but often results in blurry images due to insufficient recovery of high-frequency details. Conversely, using GAN or perceptual losses can produce sharp images with high perceptual metric scores (e.g., LPIPS), but may introduce artifacts and incorrect textures. Balancing these two types of losses can help achieve a trade-off between distortion and perception, but the challenge lies in tuning the loss function weights. To address this issue, we propose a novel method that incorporates Multi-Objective Optimization (MOO) into the training process of SISR models to balance perceptual quality and distortion. We conceptualize the relationship between loss weights and image quality assessment (IQA) metrics as black-box objective functions to be optimized within our Multi-Objective Bayesian Optimization Super-Resolution (MOBOSR) framework. This approach automates the hyperparameter tuning process, reduces overall computational cost, and enables the use of numerous loss functions simultaneously. Extensive experiments demonstrate that MOBOSR outperforms state-of-the-art methods in terms of both perceptual quality and distortion, significantly advancing the perception-distortion Pareto frontier. Our work points towards a new direction for future research on balancing perceptual quality and fidelity in nearly all image restoration tasks. The source code and pretrained models are available at: https://github.com/ZhuKeven/MOBOSR.
Authors: Zhengzhuo Xu, Bowen Qu, Yiyan Qi, Sinan Du, Chengjin Xu, Chun Yuan, Jian Guo
Abstract: Automatic chart understanding is crucial for content comprehension and document parsing. Multimodal large language models (MLLMs) have demonstrated remarkable capabilities in chart understanding through domain-specific alignment and fine-tuning. However, the application of alignment training within the chart domain is still underexplored. To address this, we propose ChartMoE, which employs the mixture of expert (MoE) architecture to replace the traditional linear projector to bridge the modality gap. Specifically, we train multiple linear connectors through distinct alignment tasks, which are utilized as the foundational initialization parameters for different experts. Additionally, we introduce ChartMoE-Align, a dataset with over 900K chart-table-JSON-code quadruples to conduct three alignment tasks (chart-table/JSON/code). Combined with the vanilla connector, we initialize different experts in four distinct ways and adopt high-quality knowledge learning to further refine the MoE connector and LLM parameters. Extensive experiments demonstrate the effectiveness of the MoE connector and our initialization strategy, e.g., ChartMoE improves the accuracy of the previous state-of-the-art from 80.48% to 84.64% on the ChartQA benchmark.
Authors: Nayeong Kim, Juwon Kang, Sungsoo Ahn, Jungseul Ok, Suha Kwak
Abstract: We study the problem of training an unbiased and accurate model given a dataset with multiple biases. This problem is challenging since the multiple biases cause multiple undesirable shortcuts during training, and even worse, mitigating one may exacerbate the other. We propose a novel training method to tackle this challenge. Our method first groups training data so that different groups induce different shortcuts, and then optimizes a linear combination of group-wise losses while adjusting their weights dynamically to alleviate conflicts between the groups in performance; this approach, rooted in the multi-objective optimization theory, encourages to achieve the minimax Pareto solution. We also present a new benchmark with multiple biases, dubbed MultiCelebA, for evaluating debiased training methods under realistic and challenging scenarios. Our method achieved the best on three datasets with multiple biases, and also showed superior performance on conventional single-bias datasets.
Authors: Junpei Honma, Akisato Kimura, Go Irie
Abstract: Measuring 3D geometric structures of indoor scenes requires dedicated depth sensors, which are not always available. Echo-based depth estimation has recently been studied as a promising alternative solution. All previous studies have assumed the use of echoes in the audible range. However, one major problem is that audible echoes cannot be used in quiet spaces or other situations where producing audible sounds is prohibited. In this paper, we consider echo-based depth estimation using inaudible ultrasonic echoes. While ultrasonic waves provide high measurement accuracy in theory, the actual depth estimation accuracy when ultrasonic echoes are used has remained unclear, due to its disadvantage of being sensitive to noise and susceptible to attenuation. We first investigate the depth estimation accuracy when the frequency of the sound source is restricted to the high-frequency band, and found that the accuracy decreased when the frequency was limited to ultrasonic ranges. Based on this observation, we propose a novel deep learning method to improve the accuracy of ultrasonic echo-based depth estimation by using audible echoes as auxiliary data only during training. Experimental results with a public dataset demonstrate that our method improves the estimation accuracy.
Authors: Shahzaib Iqbal, Tariq M. Khan, Syed S. Naqvi, Asim Naveed, Erik Meijering
Abstract: Deep learning has shown great potential for automated medical image segmentation to improve the precision and speed of disease diagnostics. However, the task presents significant difficulties due to variations in the scale, shape, texture, and contrast of the pathologies. Traditional convolutional neural network (CNN) models have certain limitations when it comes to effectively modelling multiscale context information and facilitating information interaction between skip connections across levels. To overcome these limitations, a novel deep learning architecture is introduced for medical image segmentation, taking advantage of CNNs and vision transformers. Our proposed model, named TBConvL-Net, involves a hybrid network that combines the local features of a CNN encoder-decoder architecture with long-range and temporal dependencies using biconvolutional long-short-term memory (LSTM) networks and vision transformers (ViT). This enables the model to capture contextual channel relationships in the data and account for the uncertainty of segmentation over time. Additionally, we introduce a novel composite loss function that considers both the segmentation robustness and the boundary agreement of the predicted output with the gold standard. Our proposed model shows consistent improvement over the state of the art on ten publicly available datasets of seven different medical imaging modalities.
Authors: Miaomiao Wang, Guang Hua, Sheng Li, Guorui Feng
Abstract: Virtual faces are crucial content in the metaverse. Recently, attempts have been made to generate virtual faces for privacy protection. Nevertheless, these virtual faces either permanently remove the identifiable information or map the original identity into a virtual one, which loses the original identity forever. In this study, we first attempt to address the conflict between privacy and identifiability in virtual faces, where a key-driven face anonymization and authentication recognition (KFAAR) framework is proposed. Concretely, the KFAAR framework consists of a head posture-preserving virtual face generation (HPVFG) module and a key-controllable virtual face authentication (KVFA) module. The HPVFG module uses a user key to project the latent vector of the original face into a virtual one. Then it maps the virtual vectors to obtain an extended encoding, based on which the virtual face is generated. By simultaneously adding a head posture and facial expression correction module, the virtual face has the same head posture and facial expression as the original face. During the authentication, we propose a KVFA module to directly recognize the virtual faces using the correct user key, which can obtain the original identity without exposing the original face image. We also propose a multi-task learning objective to train HPVFG and KVFA. Extensive experiments demonstrate the advantages of the proposed HPVFG and KVFA modules, which effectively achieve both facial anonymity and identifiability.
Authors: Till Nicke, Jan Raphael Schaefer, Henning Hoefener, Friedrich Feuerhake, Dorit Merhof, Fabian Kiessling, Johannes Lotz
Abstract: Due to the increasing workload of pathologists, the need for automation to support diagnostic tasks and quantitative biomarker evaluation is becoming more and more apparent. Foundation models have the potential to improve generalizability within and across centers and serve as starting points for data efficient development of specialized yet robust AI models. However, the training foundation models themselves is usually very expensive in terms of data, computation, and time. This paper proposes a supervised training method that drastically reduces these expenses. The proposed method is based on multi-task learning to train a joint encoder, by combining 16 different classification, segmentation, and detection tasks on a total of 912,000 patches. Since the encoder is capable of capturing the properties of the samples, we term it the Tissue Concepts encoder. To evaluate the performance and generalizability of the Tissue Concepts encoder across centers, classification of whole slide images from four of the most prevalent solid cancers - breast, colon, lung, and prostate - was used. The experiments show that the Tissue Concepts model achieve comparable performance to models trained with self-supervision, while requiring only 6% of the amount of training patches. Furthermore, the Tissue Concepts encoder outperforms an ImageNet pre-trained encoder on both in-domain and out-of-domain data.
Authors: Ali Aghababaei-Harandi, Massih-Reza Amini
Abstract: Despite their high accuracy, complex neural networks demand significant computational resources, posing challenges for deployment on resource-constrained devices such as mobile phones and embedded systems. Compression algorithms have been developed to address these challenges by reducing model size and computational demands while maintaining accuracy. Among these approaches, factorization methods based on tensor decomposition are theoretically sound and effective. However, they face difficulties in selecting the appropriate rank for decomposition. This paper tackles this issue by presenting a unified framework that simultaneously applies decomposition and optimal rank selection, employing a composite compression loss within defined rank constraints. Our approach includes an automatic rank search in a continuous space, efficiently identifying optimal rank configurations without the use of training data, making it computationally efficient. Combined with a subsequent fine-tuning step, our approach maintains the performance of highly compressed models on par with their original counterparts. Using various benchmark datasets, we demonstrate the efficacy of our method through a comprehensive analysis.
Authors: Philipp Quentin, Daniel Goehring
Abstract: For the use of 6D pose estimation in robotic applications, reliable poses are of utmost importance to ensure a safe, reliable and predictable operational performance. Despite these requirements, state-of-the-art 6D pose estimators often do not provide any uncertainty quantification for their pose estimates at all, or if they do, it has been shown that the uncertainty provided is only weakly correlated with the actual true error. To address this issue, we investigate a simple but effective uncertainty quantification, that we call MaskVal, which compares the pose estimates with their corresponding instance segmentations by rendering and does not require any modification of the pose estimator itself. Despite its simplicity, MaskVal significantly outperforms a state-of-the-art ensemble method on both a dataset and a robotic setup. We show that by using MaskVal, the performance of a state-of-the-art 6D pose estimator is significantly improved towards a safe and reliable operation. In addition, we propose a new and specific approach to compare and evaluate uncertainty quantification methods for 6D pose estimation in the context of robotic manipulation.
Authors: Georg Siedel, Ekagra Gupta, Andrey Morozov
Abstract: Robustness is critical for machine learning (ML) classifiers to ensure consistent performance in real-world applications where models may encounter corrupted or adversarial inputs. In particular, assessing the robustness of classifiers to adversarial inputs is essential to protect systems from vulnerabilities and thus ensure safety in use. However, methods to accurately compute adversarial robustness have been challenging for complex ML models and high-dimensional data. Furthermore, evaluations typically measure adversarial accuracy on specific attack budgets, limiting the informative value of the resulting metrics. This paper investigates the estimation of the more informative adversarial distance using iterative adversarial attacks and a certification approach. Combined, the methods provide a comprehensive evaluation of adversarial robustness by computing estimates for the upper and lower bounds of the adversarial distance. We present visualisations and ablation studies that provide insights into how this evaluation method should be applied and parameterised. We find that our adversarial attack approach is effective compared to related implementations, while the certification method falls short of expectations. The approach in this paper should encourage a more informative way of evaluating the adversarial robustness of ML classifiers.
Authors: Stephen Tian, Blake Wulfe, Kyle Sargent, Katherine Liu, Sergey Zakharov, Vitor Guizilini, Jiajun Wu
Abstract: Large-scale visuomotor policy learning is a promising approach toward developing generalizable manipulation systems. Yet, policies that can be deployed on diverse embodiments, environments, and observational modalities remain elusive. In this work, we investigate how knowledge from large-scale visual data of the world may be used to address one axis of variation for generalizable manipulation: observational viewpoint. Specifically, we study single-image novel view synthesis models, which learn 3D-aware scene-level priors by rendering images of the same scene from alternate camera viewpoints given a single input image. For practical application to diverse robotic data, these models must operate zero-shot, performing view synthesis on unseen tasks and environments. We empirically analyze view synthesis models within a simple data-augmentation scheme that we call View Synthesis Augmentation (VISTA) to understand their capabilities for learning viewpoint-invariant policies from single-viewpoint demonstration data. Upon evaluating the robustness of policies trained with our method to out-of-distribution camera viewpoints, we find that they outperform baselines in both simulated and real-world manipulation tasks. Videos and additional visualizations are available at https://s-tian.github.io/projects/vista.
Authors: Hu Wang, Congbo Ma, Jianpeng Zhang, Wei Emma Zhang, Gustavo Carneiro
Abstract: Current deep image super-resolution (SR) approaches aim to restore high-resolution images from down-sampled images or by assuming degradation from simple Gaussian kernels and additive noises. However, these techniques only assume crude approximations of the real-world image degradation process, which should involve complex kernels and noise patterns that are difficult to model using simple assumptions. In this paper, we propose a more realistic process to synthesise low-resolution images for real-world image SR by introducing a new Kernel Adversarial Learning Super-resolution (KASR) framework. In the proposed framework, degradation kernels and noises are adaptively modelled rather than explicitly specified. Moreover, we also propose a high-frequency selective objective and an iterative supervision process to further boost the model SR reconstruction accuracy. Extensive experiments validate the effectiveness of the proposed framework on real-world datasets.
Authors: Ruiping Liu, Kailun Yang, Alina Roitberg, Jiaming Zhang, Kunyu Peng, Huayao Liu, Yaonan Wang, Rainer Stiefelhagen
Abstract: Semantic segmentation benchmarks in the realm of autonomous driving are dominated by large pre-trained transformers, yet their widespread adoption is impeded by substantial computational costs and prolonged training durations. To lift this constraint, we look at efficient semantic segmentation from a perspective of comprehensive knowledge distillation and aim to bridge the gap between multi-source knowledge extractions and transformer-specific patch embeddings. We put forward the Transformer-based Knowledge Distillation (TransKD) framework which learns compact student transformers by distilling both feature maps and patch embeddings of large teacher transformers, bypassing the long pre-training process and reducing the FLOPs by >85.0%. Specifically, we propose two fundamental modules to realize feature map distillation and patch embedding distillation, respectively: (1) Cross Selective Fusion (CSF) enables knowledge transfer between cross-stage features via channel attention and feature map distillation within hierarchical transformers; (2) Patch Embedding Alignment (PEA) performs dimensional transformation within the patchifying process to facilitate the patch embedding distillation. Furthermore, we introduce two optimization modules to enhance the patch embedding distillation from different perspectives: (1) Global-Local Context Mixer (GL-Mixer) extracts both global and local information of a representative embedding; (2) Embedding Assistant (EA) acts as an embedding method to seamlessly bridge teacher and student models with the teacher's number of channels. Experiments on Cityscapes, ACDC, NYUv2, and Pascal VOC2012 datasets show that TransKD outperforms state-of-the-art distillation frameworks and rivals the time-consuming pre-training method. The source code is publicly available at https://github.com/RuipingL/TransKD.
Authors: Zihu Wang, Yu Wang, Zhuotong Chen, Hanbin Hu, Peng Li
Abstract: Contrastive learning demonstrates great promise for representation learning. Data augmentations play a critical role in contrastive learning by providing informative views of the data without necessitating explicit labels. Nonetheless, the efficacy of current methodologies heavily hinges on the quality of employed data augmentation (DA) functions, often chosen manually from a limited set of options. While exploiting diverse data augmentations is appealing, the complexities inherent in both DAs and representation learning can lead to performance deterioration. Addressing this challenge and facilitating the systematic incorporation of diverse data augmentations, this paper proposes Contrastive Learning with Consistent Representations CoCor. At the heart of CoCor is a novel consistency metric termed DA consistency. This metric governs the mapping of augmented input data to the representation space, ensuring that these instances are positioned optimally in a manner consistent with the applied intensity of the DA. Moreover, we propose to learn the optimal mapping locations as a function of DA, all while preserving a desired monotonic property relative to DA intensity. Experimental results demonstrate that CoCor notably enhances the generalizability and transferability of learned representations in comparison to baseline methods.
Authors: Pengfei Pei
Abstract: With the advancement of deep learning-driven video editing technology, security risks have emerged. Malicious video tampering can lead to public misunderstanding, property losses, and legal disputes. Currently, detection methods are mostly limited to specific datasets, with limited detection performance for unknown forgeries, and lack of robustness for processed data. This paper proposes an effective video tampering localization network that significantly improves the detection performance of video inpainting and splicing by extracting more generalized features of forgery traces. Considering the inherent differences between tampered videos and original videos, such as edge artifacts, pixel distribution, texture features, and compress information, we have specifically designed four modules to independently extract these features. Furthermore, to seamlessly integrate these features, we employ a two-stage approach utilizing both a Convolutional Neural Network and a Vision Transformer, enabling us to learn these features in a local-to-global manner. Experimental results demonstrate that the method significantly outperforms the existing state-of-the-art methods and exhibits robustness.
Authors: Xiang Feng, Chengkai Wang, Chengyu Wu, Yunxiang Li, Yongbo He, Shuai Wang, Yaiqi Wang
Abstract: Precise Tooth Cone Beam Computed Tomography (CBCT) image segmentation is crucial for orthodontic treatment planning. In this paper, we propose FDNet, a Feature Decoupled Segmentation Network, to excel in the face of the variable dental conditions encountered in CBCT scans, such as complex artifacts and indistinct tooth boundaries. The Low-Frequency Wavelet Transform (LF-Wavelet) is employed to enrich the semantic content by emphasizing the global structural integrity of the teeth, while the SAM encoder is leveraged to refine the boundary delineation, thus improving the contrast between adjacent dental structures. By integrating these dual aspects, FDNet adeptly addresses the semantic gap, providing a detailed and accurate segmentation. The framework's effectiveness is validated through rigorous benchmarks, achieving the top Dice and IoU scores of 85.28% and 75.23%, respectively. This innovative decoupling of semantic and boundary features capitalizes on the unique strengths of each element to elevate the quality of segmentation performance.
Authors: Supreeth Mysore Venkatesh, Antonio Macaluso, Marlon Nuske, Matthias Klusch, Andreas Dengel
Abstract: We present Q-Seg, a novel unsupervised image segmentation method based on quantum annealing, tailored for existing quantum hardware. We formulate the pixel-wise segmentation problem, which assimilates spectral and spatial information of the image, as a graph-cut optimization task. Our method efficiently leverages the interconnected qubit topology of the D-Wave Advantage device, offering superior scalability over existing quantum approaches and outperforming several tested state-of-the-art classical methods. Empirical evaluations on synthetic datasets have shown that Q-Seg has better runtime performance than the state-of-the-art classical optimizer Gurobi. The method has also been tested on earth observation image segmentation, a critical area with noisy and unreliable annotations. In the era of noisy intermediate-scale quantum, Q-Seg emerges as a reliable contender for real-world applications in comparison to advanced techniques like Segment Anything. Consequently, Q-Seg offers a promising solution using available quantum hardware, especially in situations constrained by limited labeled data and the need for efficient computational runtime.
Authors: Bowen Fu, Gu Wang, Chenyangguang Zhang, Yan Di, Ziqin Huang, Zhiying Leng, Fabian Manhardt, Xiangyang Ji, Federico Tombari
Abstract: Reconstructing hand-held objects from a single RGB image is a challenging task in computer vision. In contrast to prior works that utilize deterministic modeling paradigms, we employ a point cloud denoising diffusion model to account for the probabilistic nature of this problem. In the core, we introduce centroid-fixed dual-stream conditional diffusion for monocular hand-held object reconstruction (D-SCo), tackling two predominant challenges. First, to avoid the object centroid from deviating, we utilize a novel hand-constrained centroid fixing paradigm, enhancing the stability of diffusion and reverse processes and the precision of feature projection. Second, we introduce a dual-stream denoiser to semantically and geometrically model hand-object interactions with a novel unified hand-object semantic embedding, enhancing the reconstruction performance of the hand-occluded region of the object. Experiments on the synthetic ObMan dataset and three real-world datasets HO3D, MOW and DexYCB demonstrate that our approach can surpass all other state-of-the-art methods.
Authors: Renao Yan, Qiehe Sun, Cheng Jin, Yiqing Liu, Yonghong He, Tian Guan, Hao Chen
Abstract: In computational pathology, whole-slide image (WSI) classification presents a formidable challenge due to its gigapixel resolution and limited fine-grained annotations. Multiple-instance learning (MIL) offers a weakly supervised solution, yet refining instance-level information from bag-level labels remains challenging. While most of the conventional MIL methods use attention scores to estimate instance importance scores (IIS) which contribute to the prediction of the slide labels, these often lead to skewed attention distributions and inaccuracies in identifying crucial instances. To address these issues, we propose a new approach inspired by cooperative game theory: employing Shapley values to assess each instance's contribution, thereby improving IIS estimation. The computation of the Shapley value is then accelerated using attention, meanwhile retaining the enhanced instance identification and prioritization. We further introduce a framework for the progressive assignment of pseudo bags based on estimated IIS, encouraging more balanced attention distributions in MIL models. Our extensive experiments on CAMELYON-16, BRACS, TCGA-LUNG, and TCGA-BRCA datasets show our method's superiority over existing state-of-the-art approaches, offering enhanced interpretability and class-wise insights. Our source code is available at https://github.com/RenaoYan/PMIL.
Authors: Renjie Wu, Hu Wang, Feras Dayoub, Hsiang-Ting Chen
Abstract: Augmented Reality (AR) devices, emerging as prominent mobile interaction platforms, face challenges in user safety, particularly concerning oncoming vehicles. While some solutions leverage onboard camera arrays, these cameras often have limited field-of-view (FoV) with front or downward perspectives. Addressing this, we propose a new out-of-view semantic segmentation task and Segment Beyond View (SBV), a novel audio-visual semantic segmentation method. SBV supplements the visual modality, which miss the information beyond FoV, with the auditory information using a teacher-student distillation model (Omni2Ego). The model consists of a vision teacher utilising panoramic information, an auditory teacher with 8-channel audio, and an audio-visual student that takes views with limited FoV and binaural audio as input and produce semantic segmentation for objects outside FoV. SBV outperforms existing models in comparative evaluations and shows a consistent performance across varying FoV ranges and in monaural audio settings.
Authors: Zhen Zhou, Junfeng Fan, Yunkai Ma, Sihan Zhao, Fengshui Jing, Min Tan
Abstract: Instance segmentation in unmanned aerial vehicle measurement is a long-standing challenge. Since horizontal bounding boxes introduce many interference objects, oriented bounding boxes (OBBs) are usually used for instance identification. However, based on ``segmentation within bounding box'' paradigm, current instance segmentation methods using OBBs are overly dependent on bounding box detection performance. To tackle this, this paper proposes OBSeg, an efficient instance segmentation framework using OBBs. OBSeg is based on box prompt-based segmentation foundation models (BSMs), e.g., Segment Anything Model. Specifically, OBSeg first detects OBBs to distinguish instances and provide coarse localization information. Then, it predicts OBB prompt-related masks for fine segmentation. Since OBBs only serve as prompts, OBSeg alleviates the over-dependence on bounding box detection performance of current instance segmentation methods using OBBs. In addition, to enable BSMs to handle OBB prompts, we propose a novel OBB prompt encoder. To make OBSeg more lightweight and further improve the performance of lightweight distilled BSMs, a Gaussian smoothing-based knowledge distillation method is introduced. Experiments demonstrate that OBSeg outperforms current instance segmentation methods on multiple public datasets. The code is available at https://github.com/zhen6618/OBBInstanceSegmentation.
Authors: Jiyao Li, Mingze Ni, Yifei Dong, Tianqing Zhu, Wei Liu
Abstract: Recent advances in deep learning research have shown remarkable achievements across many tasks in computer vision (CV) and natural language processing (NLP). At the intersection of CV and NLP is the problem of image captioning, where the related models' robustness against adversarial attacks has not been well studied. This paper presents a novel adversarial attack strategy, AICAttack (Attention-based Image Captioning Attack), designed to attack image captioning models through subtle perturbations on images. Operating within a black-box attack scenario, our algorithm requires no access to the target model's architecture, parameters, or gradient information. We introduce an attention-based candidate selection mechanism that identifies the optimal pixels to attack, followed by a customised differential evolution method to optimise the perturbations of pixels' RGB values. We demonstrate AICAttack's effectiveness through extensive experiments on benchmark datasets against multiple victim models. The experimental results demonstrate that our method outperforms current leading-edge techniques by achieving consistently higher attack success rates.
Authors: Anita Rau, Josiah Aklilu, F. Christopher Holsinger, Serena Yeung-Levy
Abstract: Neural Radiance Fields (NeRFs) are trained to minimize the rendering loss of predicted viewpoints. However, the photometric loss often does not provide enough information to disambiguate between different possible geometries yielding the same image. Previous work has thus incorporated depth supervision during NeRF training, leveraging dense predictions from pre-trained depth networks as pseudo-ground truth. While these depth priors are assumed to be perfect once filtered for noise, in practice, their accuracy is more challenging to capture. This work proposes a novel approach to uncertainty in depth priors for NeRF supervision. Instead of using custom-trained depth or uncertainty priors, we use off-the-shelf pretrained diffusion models to predict depth and capture uncertainty during the denoising process. Because we know that depth priors are prone to errors, we propose to supervise the ray termination distance distribution with Earth Mover's Distance instead of enforcing the rendered depth to replicate the depth prior exactly through L2-loss. Our depth-guided NeRF outperforms all baselines on standard depth metrics by a large margin while maintaining performance on photometric measures.
Authors: Li Mi, Chang Xu, Javiera Castillo-Navarro, Syrielle Montariol, Wen Yang, Antoine Bosselut, Devis Tuia
Abstract: Cross-view geo-localization aims at localizing a ground-level query image by matching it to its corresponding geo-referenced aerial view. In real-world scenarios, the task requires accommodating diverse ground images captured by users with varying orientations and reduced field of views (FoVs). However, existing learning pipelines are orientation-specific or FoV-specific, demanding separate model training for different ground view variations. Such models heavily depend on the North-aligned spatial correspondence and predefined FoVs in the training data, compromising their robustness across different settings. To tackle this challenge, we propose ConGeo, a single- and cross-view Contrastive method for Geo-localization: it enhances robustness and consistency in feature representations to improve a model's invariance to orientation and its resilience to FoV variations, by enforcing proximity between ground view variations of the same location. As a generic learning objective for cross-view geo-localization, when integrated into state-of-the-art pipelines, ConGeo significantly boosts the performance of three base models on four geo-localization benchmarks for diverse ground view variations and outperforms competing methods that train separate models for each ground view variation.
Authors: Yujin Chen, Yinyu Nie, Benjamin Ummenhofer, Reiner Birkl, Michael Paulitsch, Matthias M\"uller, Matthias Nie{\ss}ner
Abstract: We present Mesh2NeRF, an approach to derive ground-truth radiance fields from textured meshes for 3D generation tasks. Many 3D generative approaches represent 3D scenes as radiance fields for training. Their ground-truth radiance fields are usually fitted from multi-view renderings from a large-scale synthetic 3D dataset, which often results in artifacts due to occlusions or under-fitting issues. In Mesh2NeRF, we propose an analytic solution to directly obtain ground-truth radiance fields from 3D meshes, characterizing the density field with an occupancy function featuring a defined surface thickness, and determining view-dependent color through a reflection function considering both the mesh and environment lighting. Mesh2NeRF extracts accurate radiance fields which provides direct supervision for training generative NeRFs and single scene representation. We validate the effectiveness of Mesh2NeRF across various tasks, achieving a noteworthy 3.12dB improvement in PSNR for view synthesis in single scene representation on the ABO dataset, a 0.69 PSNR enhancement in the single-view conditional generation of ShapeNet Cars, and notably improved mesh extraction from NeRF in the unconditional generation of Objaverse Mugs.
Authors: Yanglin Feng, Yang Qin, Dezhong Peng, Hongyuan Zhu, Xi Peng, Peng Hu
Abstract: In this paper, we present and study a new instance-level retrieval task: PointCloud-Text Matching~(PTM), which aims to find the exact cross-modal instance that matches a given point-cloud query or text query. PTM could be applied to various scenarios, such as indoor/urban-canyon localization and scene retrieval. However, there exists no suitable and targeted dataset for PTM in practice. Therefore, we construct three new PTM benchmark datasets, namely 3D2T-SR, 3D2T-NR, and 3D2T-QA. We observe that the data is challenging and with noisy correspondence due to the sparsity, noise, or disorder of point clouds and the ambiguity, vagueness, or incompleteness of texts, which make existing cross-modal matching methods ineffective for PTM. To tackle these challenges, we propose a PTM baseline, named Robust PointCloud-Text Matching method (RoMa). RoMa consists of two modules: a Dual Attention Perception module (DAP) and a Robust Negative Contrastive Learning module (RNCL). Specifically, DAP leverages token-level and feature-level attention to adaptively focus on useful local and global features, and aggregate them into common representations, thereby reducing the adverse impact of noise and ambiguity. To handle noisy correspondence, RNCL divides negative pairs, which are much less error-prone than positive pairs, into clean and noisy subsets, and assigns them forward and reverse optimization directions respectively, thus enhancing robustness against noisy correspondence. We conduct extensive experiments on our benchmarks and demonstrate the superiority of our RoMa.
Authors: Linhui Xiao, Xiaoshan Yang, Fang Peng, Yaowei Wang, Changsheng Xu
Abstract: Visual grounding, which aims to ground a visual region via natural language, is a task that heavily relies on cross-modal alignment. Existing works utilized uni-modal pre-trained models to transfer visual or linguistic knowledge separately while ignoring the multimodal corresponding information. Motivated by recent advancements in contrastive language-image pre-training and low-rank adaptation (LoRA) methods, we aim to solve the grounding task based on multimodal pre-training. However, there exists significant task gaps between pre-training and grounding. Therefore, to address these gaps, we propose a concise and efficient hierarchical multimodal fine-grained modulation framework, namely HiVG. Specifically, HiVG consists of a multi-layer adaptive cross-modal bridge and a hierarchical multimodal low-rank adaptation (HiLoRA) paradigm. The cross-modal bridge can address the inconsistency between visual features and those required for grounding, and establish a connection between multi-level visual and text features. HiLoRA prevents the accumulation of perceptual errors by adapting the cross-modal features from shallow to deep layers in a hierarchical manner. Experimental results on five datasets demonstrate the effectiveness of our approach and showcase the significant grounding capabilities as well as promising energy efficiency advantages. The project page: https://github.com/linhuixiao/HiVG.
Authors: Bin Wang, Zhuangcheng Gu, Guang Liang, Chao Xu, Bo Zhang, Botian Shi, Conghui He
Abstract: The paper introduces the UniMER dataset, marking the first study on Mathematical Expression Recognition (MER) targeting complex real-world scenarios. The UniMER dataset includes a large-scale training set, UniMER-1M, which offers unprecedented scale and diversity with one million training instances to train high-quality, robust models. Additionally, UniMER features a meticulously designed, diverse test set, UniMER-Test, which covers a variety of formula distributions found in real-world scenarios, providing a more comprehensive and fair evaluation. To better utilize the UniMER dataset, the paper proposes a Universal Mathematical Expression Recognition Network (UniMERNet), tailored to the characteristics of formula recognition. UniMERNet consists of a carefully designed encoder that incorporates detail-aware and local context features, and an optimized decoder for accelerated performance. Extensive experiments conducted using the UniMER-1M dataset and UniMERNet demonstrate that training on the large-scale UniMER-1M dataset can produce a more generalizable formula recognition model, significantly outperforming all previous datasets. Furthermore, the introduction of UniMERNet enhances the model's performance in formula recognition, achieving higher accuracy and speeds. All data, models, and code are available at https://github.com/opendatalab/UniMERNet.
Authors: Junyi Ma, Jingyi Xu, Xieyuanli Chen, Hesheng Wang
Abstract: Understanding how humans would behave during hand-object interaction is vital for applications in service robot manipulation and extended reality. To achieve this, some recent works have been proposed to simultaneously forecast hand trajectories and object affordances on human egocentric videos. The joint prediction serves as a comprehensive representation of future hand-object interactions in 2D space, indicating potential human motion and motivation. However, the existing approaches mostly adopt the autoregressive paradigm for unidirectional prediction, which lacks mutual constraints within the holistic future sequence, and accumulates errors along the time axis. Meanwhile, these works basically overlook the effect of camera egomotion on first-person view predictions. To address these limitations, we propose a novel diffusion-based interaction prediction method, namely Diff-IP2D, to forecast future hand trajectories and object affordances concurrently in an iterative non-autoregressive manner. We transform the sequential 2D images into latent feature space and design a denoising diffusion model to predict future latent interaction features conditioned on past ones. Motion features are further integrated into the conditional denoising process to enable Diff-IP2D aware of the camera wearer's dynamics for more accurate interaction prediction. Extensive experiments demonstrate that our method significantly outperforms the state-of-the-art baselines on both the off-the-shelf metrics and our newly proposed evaluation protocol. This highlights the efficacy of leveraging a generative paradigm for 2D hand-object interaction prediction. The code of Diff-IP2D will be released at https://github.com/IRMVLab/Diff-IP2D.
Authors: Hanchen Tai, Qingdong He, Jiangning Zhang, Yijie Qian, Zhenyu Zhang, Xiaobin Hu, Xiangtai Li, Yabiao Wang, Yong Liu
Abstract: Open-vocabulary 3D scene understanding presents a significant challenge in the field. Recent works have sought to transfer knowledge embedded in vision-language models from 2D to 3D domains. However, these approaches often require prior knowledge from specific 3D scene datasets, limiting their applicability in open-world scenarios. The Segment Anything Model (SAM) has demonstrated remarkable zero-shot segmentation capabilities, prompting us to investigate its potential for comprehending 3D scenes without training. In this paper, we introduce OV-SAM3D, a training-free method that contains a universal framework for understanding open-vocabulary 3D scenes. This framework is designed to perform understanding tasks for any 3D scene without requiring prior knowledge of the scene. Specifically, our method is composed of two key sub-modules: First, we initiate the process by generating superpoints as the initial 3D prompts and refine these prompts using segment masks derived from SAM. Moreover, we then integrate a specially designed overlapping score table with open tags from the Recognize Anything Model (RAM) to produce final 3D instances with open-world labels. Empirical evaluations on the ScanNet200 and nuScenes datasets demonstrate that our approach surpasses existing open-vocabulary methods in unknown open-world environments.
Authors: Maria Korosteleva, Timur Levent Kesdogan, Fabian Kemper, Stephan Wenninger, Jasmin Koller, Yuhan Zhang, Mario Botsch, Olga Sorkine-Hornung
Abstract: Recent research interest in the learning-based processing of garments, from virtual fitting to generation and reconstruction, stumbles on a scarcity of high-quality public data in the domain. We contribute to resolving this need by presenting the first large-scale synthetic dataset of 3D made-to-measure garments with sewing patterns, as well as its generation pipeline. GarmentCodeData contains 115,000 data points that cover a variety of designs in many common garment categories: tops, shirts, dresses, jumpsuits, skirts, pants, etc., fitted to a variety of body shapes sampled from a custom statistical body model based on CAESAR, as well as a standard reference body shape, applying three different textile materials. To enable the creation of datasets of such complexity, we introduce a set of algorithms for automatically taking tailor's measures on sampled body shapes, sampling strategies for sewing pattern design, and propose an automatic, open-source 3D garment draping pipeline based on a fast XPBD simulator, while contributing several solutions for collision resolution and drape correctness to enable scalability. Project Page: https://igl.ethz.ch/projects/GarmentCodeData/
Authors: Angel Villar-Corrales, Moritz Austermann, Sven Behnke
Abstract: Autonomous systems, such as self-driving cars, rely on reliable semantic environment perception for decision making. Despite great advances in video semantic segmentation, existing approaches ignore important inductive biases and lack structured and interpretable internal representations. In this work, we propose MCDS-VSS, a structured filter model that learns in a self-supervised manner to estimate scene geometry and ego-motion of the camera, while also estimating the motion of external objects. Our model leverages these representations to improve the temporal consistency of semantic segmentation without sacrificing segmentation accuracy. MCDS-VSS follows a prediction-fusion approach in which scene geometry and camera motion are first used to compensate for ego-motion, then residual flow is used to compensate motion of dynamic objects, and finally the predicted scene features are fused with the current features to obtain a temporally consistent scene segmentation. Our model parses automotive scenes into multiple decoupled interpretable representations such as scene geometry, ego-motion, and object motion. Quantitative evaluation shows that MCDS-VSS achieves superior temporal consistency on video sequences while retaining competitive segmentation performance.
Authors: Richa Upadhyay, Ronald Phlypo, Rajkumar Saini, Marcus Liwicki
Abstract: In the Multi-task Learning (MTL) framework, every task demands distinct feature representations, ranging from low-level to high-level attributes. It is vital to address the specific (feature/parameter) needs of each task, especially in computationally constrained environments. This work, therefore, introduces Layer-Optimized Multi-Task (LOMT) models that utilize structured sparsity to refine feature selection for individual tasks and enhance the performance of all tasks in a multi-task scenario. Structured or group sparsity systematically eliminates parameters from trivial channels and, sometimes, eventually, entire layers within a convolution neural network during training. Consequently, the remaining layers provide the most optimal features for a given task. In this two-step approach, we subsequently leverage this sparsity-induced optimal layer information to build the LOMT models by connecting task-specific decoders to these strategically identified layers, deviating from conventional approaches that uniformly connect decoders at the end of the network. This tailored architecture optimizes the network, focusing on essential features while reducing redundancy. We validate the efficacy of the proposed approach on two datasets, i.e., NYU-v2 and CelebAMask-HD datasets, for multiple heterogeneous tasks. A detailed performance analysis of the LOMT models, in contrast to the conventional MTL models, reveals that the LOMT models outperform for most task combinations. The excellent qualitative and quantitative outcomes highlight the effectiveness of employing structured sparsity for optimal layer (or feature) selection.
Authors: Weiwei Duan, Luping Ji, Shengjia Chen, Sicheng Zhu, Mao Ye
Abstract: As a sub-field of object detection, moving infrared small target detection presents significant challenges due to tiny target sizes and low contrast against backgrounds. Currently-existing methods primarily rely on the features extracted only from spatio-temporal domain. Frequency domain has hardly been concerned yet, although it has been widely applied in image processing. To extend feature source domains and enhance feature representation, we propose a new Triple-domain Strategy (Tridos) with the frequency-aware memory enhancement on spatio-temporal domain for infrared small target detection. In this scheme, it effectively detaches and enhances frequency features by a local-global frequency-aware module with Fourier transform. Inspired by human visual system, our memory enhancement is designed to capture the spatial relations of infrared targets among video frames. Furthermore, it encodes temporal dynamics motion features via differential learning and residual enhancing. Additionally, we further design a residual compensation to reconcile possible cross-domain feature mismatches. To our best knowledge, proposed Tridos is the first work to explore infrared target feature learning comprehensively in spatio-temporal-frequency domains. The extensive experiments on three datasets (i.e., DAUB, ITSDT-15K and IRDST) validate that our triple-domain infrared feature learning scheme could often be obviously superior to state-of-the-art ones. Source codes are available at https://github.com/UESTC-nnLab/Tridos.
Authors: Jiayi Liu, Qianyu Zhang, Xue Wan, Shengyang Zhang, Yaolin Tian, Haodong Han, Yutao Zhao, Baichuan Liu, Zeyuan Zhao, Xubo Luo
Abstract: With the complexity of lunar exploration missions, the moon needs to have a higher level of autonomy. Environmental perception and navigation algorithms are the foundation for lunar rovers to achieve autonomous exploration. The development and verification of algorithms require highly reliable data support. Most of the existing lunar datasets are targeted at a single task, lacking diverse scenes and high-precision ground truth labels. To address this issue, we propose a multi-task, multi-scene, and multi-label lunar benchmark dataset LuSNAR. This dataset can be used for comprehensive evaluation of autonomous perception and navigation systems, including high-resolution stereo image pairs, panoramic semantic labels, dense depth maps, LiDAR point clouds, and the position of rover. In order to provide richer scene data, we built 9 lunar simulation scenes based on Unreal Engine. Each scene is divided according to topographic relief and the density of objects. To verify the usability of the dataset, we evaluated and analyzed the algorithms of semantic segmentation, 3D reconstruction, and autonomous navigation. The experiment results prove that the dataset proposed in this paper can be used for ground verification of tasks such as autonomous environment perception and navigation, and provides a lunar benchmark dataset for testing the accessibility of algorithm metrics. We make LuSNAR publicly available at: https://github.com/autumn999999/LuSNAR-dataset.
Authors: Minghui Wu, Chenxu Zhao, Anyang Su, Donglin Di, Tianyu Fu, Da An, Min He, Ya Gao, Meng Ma, Kun Yan, Ping Wang
Abstract: Understanding of video creativity and content often varies among individuals, with differences in focal points and cognitive levels across different ages, experiences, and genders. There is currently a lack of research in this area, and most existing benchmarks suffer from several drawbacks: 1) a limited number of modalities and answers with restrictive length; 2) the content and scenarios within the videos are excessively monotonous, transmitting allegories and emotions that are overly simplistic. To bridge the gap to real-world applications, we introduce a large-scale Subjective Response Indicators for Advertisement Videos dataset, namely SRI-ADV. Specifically, we collected real changes in Electroencephalographic (EEG) and eye-tracking regions from different demographics while they viewed identical video content. Utilizing this multi-modal dataset, we developed tasks and protocols to analyze and evaluate the extent of cognitive understanding of video content among different users. Along with the dataset, we designed a Hypergraph Multi-modal Large Language Model (HMLLM) to explore the associations among different demographics, video elements, EEG, and eye-tracking indicators. HMLLM could bridge semantic gaps across rich modalities and integrate information beyond different modalities to perform logical reasoning. Extensive experimental evaluations on SRI-ADV and other additional video-based generative performance benchmarks demonstrate the effectiveness of our method. The codes and dataset will be released at https://github.com/mininglamp-MLLM/HMLLM.
Authors: Yunfeng Diao, Baiqi Wu, Ruixuan Zhang, Xun Yang, Meng Wang, He Wang
Abstract: Skeletal motion plays a pivotal role in human activity recognition (HAR). Recently, attack methods have been proposed to identify the universal vulnerability of skeleton-based HAR(S-HAR). However, the research of adversarial transferability on S-HAR is largely missing. More importantly, existing attacks all struggle in transfer across unknown S-HAR models. We observed that the key reason is that the loss landscape of the action recognizers is rugged and sharp. Given the established correlation in prior studies~\cite{qin2022boosting,wu2020towards} between loss landscape and adversarial transferability, we assume and empirically validate that smoothing the loss landscape could potentially improve adversarial transferability on S-HAR. This is achieved by proposing a new post-train Dual Bayesian strategy, which can effectively explore the model posterior space for a collection of surrogates without the need for re-training. Furthermore, to craft adversarial examples along the motion manifold, we incorporate the attack gradient with information of the motion dynamics in a Bayesian manner. Evaluated on benchmark datasets, e.g. HDM05 and NTU 60, the average transfer success rate can reach as high as 35.9\% and 45.5\% respectively. In comparison, current state-of-the-art skeletal attacks achieve only 3.6\% and 9.8\%. The high adversarial transferability remains consistent across various surrogate, victim, and even defense models. Through a comprehensive analysis of the results, we provide insights on what surrogates are more likely to exhibit transferability, to shed light on future research.
Authors: Josep Cabacas-Maso, Elena Ortega-Beltr\'an, Ismael Benito-Altamirano, Carles Ventura
Abstract: We present our contribution to the 7th ABAW challenge at ECCV 2024, by utilizing a Dual-Direction Attention Mixed Feature Network (DDAMFN) for multitask facial expression recognition, we achieve results far beyond the proposed baseline for the Multi-Task ABAW challenge. Our proposal uses the well-known DDAMFN architecture as base to effectively predict valence-arousal, emotion recognition, and facial action units. We demonstrate the architecture ability to handle these tasks simultaneously, providing insights into its architecture and the rationale behind its design. Additionally, we compare our results for a multitask solution with independent single-task performance.
Authors: Yuchen Weng, Zhengwen Shen, Ruofan Chen, Qi Wang, Jun Wang
Abstract: 3D deblurring reconstruction techniques have recently seen significant advancements with the development of Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS). Although these techniques can recover relatively clear 3D reconstructions from blurry image inputs, they still face limitations in handling severe blurring and complex camera motion. To address these issues, we propose Event-assisted 3D Deblur Reconstruction with Gaussian Splatting (EaDeblur-GS), which integrates event camera data to enhance the robustness of 3DGS against motion blur. By employing an Adaptive Deviation Estimator (ADE) network to estimate Gaussian center deviations and using novel loss functions, EaDeblur-GS achieves sharp 3D reconstructions in real-time, demonstrating performance comparable to state-of-the-art methods.
Authors: Qiao Wu, Kun Sun, Pei An, Mathieu Salzmann, Yanning Zhang, Jiaqi Yang
Abstract: The high temporal variation of the point clouds is the key challenge of 3D single-object tracking (3D SOT). Existing approaches rely on the assumption that the shape variation of the point clouds and the motion of the objects across neighboring frames are smooth, failing to cope with high temporal variation data. In this paper, we present a novel framework for 3D SOT in point clouds with high temporal variation, called HVTrack. HVTrack proposes three novel components to tackle the challenges in the high temporal variation scenario: 1) A Relative-Pose-Aware Memory module to handle temporal point cloud shape variations; 2) a Base-Expansion Feature Cross-Attention module to deal with similar object distractions in expanded search areas; 3) a Contextual Point Guided Self-Attention module for suppressing heavy background noise. We construct a dataset with high temporal variation (KITTI-HV) by setting different frame intervals for sampling in the KITTI dataset. On the KITTI-HV with 5 frame intervals, our HVTrack surpasses the state-of-the-art tracker CXTracker by 11.3%/15.7% in Success/Precision.
Authors: Yuan Tang, Xu Han, Xianzhi Li, Qiao Yu, Jinfeng Xu, Yixue Hao, Long Hu, Min Chen
Abstract: Enabling Large Language Models (LLMs) to comprehend the 3D physical world remains a significant challenge. Due to the lack of large-scale 3D-text pair datasets, the success of LLMs has yet to be replicated in 3D understanding. In this paper, we rethink this issue and propose a new task: 3D Data-Efficient Point-Language Understanding. The goal is to enable LLMs to achieve robust 3D object understanding with minimal 3D point cloud and text data pairs. To address this task, we introduce GreenPLM, which leverages more text data to compensate for the lack of 3D data. First, inspired by using CLIP to align images and text, we utilize a pre-trained point cloud-text encoder to map the 3D point cloud space to the text space. This mapping leaves us to seamlessly connect the text space with LLMs. Once the point-text-LLM connection is established, we further enhance text-LLM alignment by expanding the intermediate text space, thereby reducing the reliance on 3D point cloud data. Specifically, we generate 6M free-text descriptions of 3D objects, and design a three-stage training strategy to help LLMs better explore the intrinsic connections between different modalities. To achieve efficient modality alignment, we design a zero-parameter cross-attention module for token pooling. Extensive experimental results show that GreenPLM requires only 12% of the 3D training data used by existing state-of-the-art models to achieve superior 3D understanding. Remarkably, GreenPLM also achieves competitive performance using text-only data. The code and weights are available at: https://github.com/TangYuan96/GreenPLM.
Authors: Manuel Alejandro Diaz-Zapata (CHROMA), Wenqian Liu (CHROMA, UGA), Robin Baruffa (CHROMA), Christian Laugier (CHROMA)
Abstract: Current research in semantic bird's-eye view segmentation for autonomous driving focuses solely on optimizing neural network models using a single dataset, typically nuScenes. This practice leads to the development of highly specialized models that may fail when faced with different environments or sensor setups, a problem known as domain shift. In this paper, we conduct a comprehensive cross-dataset evaluation of state-of-the-art BEV segmentation models to assess their performance across different training and testing datasets and setups, as well as different semantic categories. We investigate the influence of different sensors, such as cameras and LiDAR, on the models' ability to generalize to diverse conditions and scenarios. Additionally, we conduct multi-dataset training experiments that improve models' BEV segmentation performance compared to single-dataset training. Our work addresses the gap in evaluating BEV segmentation models under cross-dataset validation. And our findings underscore the importance of enhancing model generalizability and adaptability to ensure more robust and reliable BEV segmentation approaches for autonomous driving applications. The code for this paper available at https://github.com/manueldiaz96/beval .
Authors: Bonan Liu, Handi Yin, Manuel Kaufmann, Jinhao He, Sammy Christen, Jie Song, Pan Hui
Abstract: We present EgoHDM, an online egocentric-inertial human motion capture (mocap), localization, and dense mapping system. Our system uses 6 inertial measurement units (IMUs) and a commodity head-mounted RGB camera. EgoHDM is the first human mocap system that offers dense scene mapping in near real-time. Further, it is fast and robust to initialize and fully closes the loop between physically plausible map-aware global human motion estimation and mocap-aware 3D scene reconstruction. Our key idea is integrating camera localization and mapping information with inertial human motion capture bidirectionally in our system. To achieve this, we design a tightly coupled mocap-aware dense bundle adjustment and physics-based body pose correction module leveraging a local body-centric elevation map. The latter introduces a novel terrain-aware contact PD controller, which enables characters to physically contact the given local elevation map thereby reducing human floating or penetration. We demonstrate the performance of our system on established synthetic and real-world benchmarks. The results show that our method reduces human localization, camera pose, and mapping accuracy error by 41%, 71%, 46%, respectively, compared to the state of the art. Our qualitative evaluations on newly captured data further demonstrate that EgoHDM can cover challenging scenarios in non-flat terrain including stepping over stairs and outdoor scenes in the wild.
Authors: Hyungkeun Park, Jong-Seok Lee
Abstract: Logit-based knowledge distillation (KD) for classification is cost-efficient compared to feature-based KD but often subject to inferior performance. Recently, it was shown that the performance of logit-based KD can be improved by effectively delivering the probability distribution for the non-target classes from the teacher model, which is known as `implicit (dark) knowledge', to the student model. Through gradient analysis, we first show that this actually has an effect of adaptively controlling the learning of implicit knowledge. Then, we propose a new loss that enables the student to learn explicit knowledge (i.e., the teacher's confidence about the target class) along with implicit knowledge in an adaptive manner. Furthermore, we propose to separate the classification and distillation tasks for effective distillation and inter-class relationship modeling. Experimental results demonstrate that the proposed method, called adaptive explicit knowledge transfer (AEKT) method, achieves improved performance compared to the state-of-the-art KD methods on the CIFAR-100 and ImageNet datasets.
Authors: Gaojie Lin, Jianwen Jiang, Chao Liang, Tianyun Zhong, Jiaqi Yang, Yanbo Zheng
Abstract: Diffusion-based video generation technology has advanced significantly, catalyzing a proliferation of research in human animation. However, the majority of these studies are confined to same-modality driving settings, with cross-modality human body animation remaining relatively underexplored. In this paper, we introduce, an end-to-end audio-driven human animation framework that ensures hand integrity, identity consistency, and natural motion. The key design of CyberHost is the Region Codebook Attention mechanism, which improves the generation quality of facial and hand animations by integrating fine-grained local features with learned motion pattern priors. Furthermore, we have developed a suite of human-prior-guided training strategies, including body movement map, hand clarity score, pose-aligned reference feature, and local enhancement supervision, to improve synthesis results. To our knowledge, CyberHost is the first end-to-end audio-driven human diffusion model capable of facilitating zero-shot video generation within the scope of human body. Extensive experiments demonstrate that CyberHost surpasses previous works in both quantitative and qualitative aspects.
Authors: Hu Wang, David Butler, Yuan Zhang, Jodie Avery, Steven Knox, Congbo Ma, Louise Hull, Gustavo Carneiro
Abstract: Endometriosis, affecting about 10\% of individuals assigned female at birth, is challenging to diagnose and manage. Diagnosis typically involves the identification of various signs of the disease using either laparoscopic surgery or the analysis of T1/T2 MRI images, with the latter being quicker and cheaper but less accurate. A key diagnostic sign of endometriosis is the obliteration of the Pouch of Douglas (POD). However, even experienced clinicians struggle with accurately classifying POD obliteration from MRI images, which complicates the training of reliable AI models. In this paper, we introduce the \underline{H}uman-\underline{AI} \underline{Co}llaborative \underline{M}ulti-modal \underline{M}ulti-rater Learning (HAICOMM) methodology to address the challenge above. HAICOMM is the first method that explores three important aspects of this problem: 1) multi-rater learning to extract a cleaner label from the multiple ``noisy'' labels available per training sample; 2) multi-modal learning to leverage the presence of T1/T2 MRI images for training and testing; and 3) human-AI collaboration to build a system that leverages the predictions from clinicians and the AI model to provide more accurate classification than standalone clinicians and AI models. Presenting results on the multi-rater T1/T2 MRI endometriosis dataset that we collected to validate our methodology, the proposed HAICOMM model outperforms an ensemble of clinicians, noisy-label learning models, and multi-rater learning methods.
Authors: Songhua Liu, Weihao Yu, Zhenxiong Tan, Xinchao Wang
Abstract: Modern diffusion models, particularly those utilizing a Transformer-based UNet for denoising, rely heavily on self-attention operations to manage complex spatial relationships, thus achieving impressive generation performance. However, this existing paradigm faces significant challenges in generating high-resolution visual content due to its quadratic time and memory complexity with respect to the number of spatial tokens. To address this limitation, we aim at a novel linear attention mechanism as an alternative in this paper. Specifically, we begin our exploration from recently introduced models with linear complexity, e.g., Mamba2, RWKV6, Gated Linear Attention, etc, and identify two key features-attention normalization and non-causal inference-that enhance high-resolution visual generation performance. Building on these insights, we introduce a generalized linear attention paradigm, which serves as a low-rank approximation of a wide spectrum of popular linear token mixers. To save the training cost and better leverage pre-trained models, we initialize our models and distill the knowledge from pre-trained StableDiffusion (SD). We find that the distilled model, termed LinFusion, achieves performance on par with or superior to the original SD after only modest training, while significantly reducing time and memory complexity. Extensive experiments on SD-v1.5, SD-v2.1, and SD-XL demonstrate that LinFusion delivers satisfactory zero-shot cross-resolution generation performance, generating high-resolution images like 16K resolution. Moreover, it is highly compatible with pre-trained SD components, such as ControlNet and IP-Adapter, requiring no adaptation efforts. Codes are available at https://github.com/Huage001/LinFusion.
Authors: Jianwen Jiang, Chao Liang, Jiaqi Yang, Gaojie Lin, Tianyun Zhong, Yanbo Zheng
Abstract: With the introduction of diffusion-based video generation techniques, audio-conditioned human video generation has recently achieved significant breakthroughs in both the naturalness of motion and the synthesis of portrait details. Due to the limited control of audio signals in driving human motion, existing methods often add auxiliary spatial signals to stabilize movements, which may compromise the naturalness and freedom of motion. In this paper, we propose an end-to-end audio-only conditioned video diffusion model named Loopy. Specifically, we designed an inter- and intra-clip temporal module and an audio-to-latents module, enabling the model to leverage long-term motion information from the data to learn natural motion patterns and improving audio-portrait movement correlation. This method removes the need for manually specified spatial motion templates used in existing methods to constrain motion during inference. Extensive experiments show that Loopy outperforms recent audio-driven portrait diffusion models, delivering more lifelike and high-quality results across various scenarios.
Authors: Haoxiang Gao, Zhongruo Wang, Yaqian Li, Kaiwen Long, Ming Yang, Yiqing Shen
Abstract: The advent of foundation models has revolutionized the fields of natural language processing and computer vision, paving the way for their application in autonomous driving (AD). This survey presents a comprehensive review of more than 40 research papers, demonstrating the role of foundation models in enhancing AD. Large language models contribute to planning and simulation in AD, particularly through their proficiency in reasoning, code generation and translation. In parallel, vision foundation models are increasingly adapted for critical tasks such as 3D object detection and tracking, as well as creating realistic driving scenarios for simulation and testing. Multi-modal foundation models, integrating diverse inputs, exhibit exceptional visual understanding and spatial reasoning, crucial for end-to-end AD. This survey not only provides a structured taxonomy, categorizing foundation models based on their modalities and functionalities within the AD domain but also delves into the methods employed in current research. It identifies the gaps between existing foundation models and cutting-edge AD approaches, thereby charting future research directions and proposing a roadmap for bridging these gaps.
Authors: Mohammad Nazeri, Junzhe Wang, Amirreza Payandeh, Xuesu Xiao
Abstract: Humans excel at efficiently navigating through crowds without collision by focusing on specific visual regions relevant to navigation. However, most robotic visual navigation methods rely on deep learning models pre-trained on vision tasks, which prioritize salient objects -- not necessarily relevant to navigation and potentially misleading. Alternative approaches train specialized navigation models from scratch, requiring significant computation. On the other hand, self-supervised learning has revolutionized computer vision and natural language processing, but its application to robotic navigation remains underexplored due to the difficulty of defining effective self-supervision signals. Motivated by these observations, in this work, we propose a Self-Supervised Vision-Action Model for Visual Navigation Pre-Training (VANP). Instead of detecting salient objects that are beneficial for tasks such as classification or detection, VANP learns to focus only on specific visual regions that are relevant to the navigation task. To achieve this, VANP uses a history of visual observations, future actions, and a goal image for self-supervision, and embeds them using two small Transformer Encoders. Then, VANP maximizes the information between the embeddings by using a mutual information maximization objective function. We demonstrate that most VANP-extracted features match with human navigation intuition. VANP achieves comparable performance as models learned end-to-end with half the training time and models trained on a large-scale, fully supervised dataset, i.e., ImageNet, with only 0.08% data.
Authors: Shubhadip Dasgupta, Satwik Pate, Divya Rathore, L. G. Divyanth, Ayan Das, Anshuman Nayak, Subhadip Dey, Asim Biswas, David C. Weindorf, Bin Li, Sergio Henrique Godinho Silva, Bruno Teixeira Ribeiro, Sanjay Srivastava, Somsubhra Chakraborty
Abstract: This study investigated the use of portable X-ray fluorescence (PXRF) spectrometry and soil image analysis for rapid soil fertility assessment, with a focus on key indicators such as available boron (B), organic carbon (OC), available manganese (Mn), available sulfur (S), and the sulfur availability index (SAI). A total of 1,133 soil samples from diverse agro-climatic zones in Eastern India were analyzed. The research integrated color and texture features from microscopic soil images, PXRF data, and auxiliary soil variables (AVs) using a Random Forest model. Results showed that combining image features (IFs) with AVs significantly improved prediction accuracy for available B (R2 = 0.80) and OC (R2 = 0.88). A data fusion approach, incorporating IFs, AVs, and PXRF data, further enhanced predictions for available Mn and SAI, with R2 values of 0.72 and 0.70, respectively. The study highlights the potential of integrating these technologies to offer rapid, cost-effective soil testing methods, paving the way for more advanced predictive models and a deeper understanding of soil fertility. Future work should explore the application of deep learning models on a larger dataset, incorporating soils from a wider range of agro-climatic zones under field conditions.
Authors: Yingxuan Li, Ryota Hinami, Kiyoharu Aizawa, Yusuke Matsui
Abstract: Recognizing characters and predicting speakers of dialogue are critical for comic processing tasks, such as voice generation or translation. However, because characters vary by comic title, supervised learning approaches like training character classifiers which require specific annotations for each comic title are infeasible. This motivates us to propose a novel zero-shot approach, allowing machines to identify characters and predict speaker names based solely on unannotated comic images. In spite of their importance in real-world applications, these task have largely remained unexplored due to challenges in story comprehension and multimodal integration. Recent large language models (LLMs) have shown great capability for text understanding and reasoning, while their application to multimodal content analysis is still an open problem. To address this problem, we propose an iterative multimodal framework, the first to employ multimodal information for both character identification and speaker prediction tasks. Our experiments demonstrate the effectiveness of the proposed framework, establishing a robust baseline for these tasks. Furthermore, since our method requires no training data or annotations, it can be used as-is on any comic series.
Authors: Enneng Yang, Li Shen, Guibing Guo, Xingwei Wang, Xiaochun Cao, Jie Zhang, Dacheng Tao
Abstract: Model merging is an efficient empowerment technique in the machine learning community that does not require the collection of raw training data and does not require expensive computation. As model merging becomes increasingly prevalent across various fields, it is crucial to understand the available model merging techniques comprehensively. However, there is a significant gap in the literature regarding a systematic and thorough review of these techniques. This survey provides a comprehensive overview of model merging methods and theories, their applications in various domains and settings, and future research directions. Specifically, we first propose a new taxonomic approach that exhaustively discusses existing model merging methods. Secondly, we discuss the application of model merging techniques in large language models, multimodal large language models, and 10+ machine learning subfields, including continual learning, multi-task learning, few-shot learning, etc. Finally, we highlight the remaining challenges of model merging and discuss future research directions. A comprehensive list of papers about model merging is available at \url{https://github.com/EnnengYang/Awesome-Model-Merging-Methods-Theories-Applications}.
URLs: https://github.com/EnnengYang/Awesome-Model-Merging-Methods-Theories-Applications
Authors: \'Aron Samuel Kov\'acs, Pedro Hermosilla, Renata G. Raidou
Abstract: We introduce G-Style, a novel algorithm designed to transfer the style of an image onto a 3D scene represented using Gaussian Splatting. Gaussian Splatting is a powerful 3D representation for novel view synthesis, as -- compared to other approaches based on Neural Radiance Fields -- it provides fast scene renderings and user control over the scene. Recent pre-prints have demonstrated that the style of Gaussian Splatting scenes can be modified using an image exemplar. However, since the scene geometry remains fixed during the stylization process, current solutions fall short of producing satisfactory results. Our algorithm aims to address these limitations by following a three-step process: In a pre-processing step, we remove undesirable Gaussians with large projection areas or highly elongated shapes. Subsequently, we combine several losses carefully designed to preserve different scales of the style in the image, while maintaining as much as possible the integrity of the original scene content. During the stylization process and following the original design of Gaussian Splatting, we split Gaussians where additional detail is necessary within our scene by tracking the gradient of the stylized color. Our experiments demonstrate that G-Style generates high-quality stylizations within just a few minutes, outperforming existing methods both qualitatively and quantitatively.