Authors: Anna Wirth-Singh, Jinlin Xiang, Minho Choi, Johannes E. Fr\"och, Luocheng Huang, Shane Colburn, Eli Shlizerman, Arka Majumdar
Abstract: Optical and hybrid convolutional neural networks (CNNs) recently have become of increasing interest to achieve low-latency, low-power image classification and computer vision tasks. However, implementing optical nonlinearity is challenging, and omitting the nonlinear layers in a standard CNN comes at a significant reduction in accuracy. In this work, we use knowledge distillation to compress modified AlexNet to a single linear convolutional layer and an electronic backend (two fully connected layers). We obtain comparable performance to a purely electronic CNN with five convolutional layers and three fully connected layers. We implement the convolution optically via engineering the point spread function of an inverse-designed meta-optic. Using this hybrid approach, we estimate a reduction in multiply-accumulate operations from 688M in a conventional electronic modified AlexNet to only 86K in the hybrid compressed network enabled by the optical frontend. This constitutes a four orders of magnitude reduction in latency and power consumption. Furthermore, we experimentally demonstrate that the classification accuracy of the system exceeds 93\% on the MNIST dataset.
Authors: Mu Wang
Abstract: The problem of Domain Adaptive in the field of Object Detection involves the transfer of object detection models from labeled source domains to unannotated target domains. Recent advancements in this field aim to address domain discrepancies by aligning pixel-pairs across domains within a non-Euclidean graphical space, thereby minimizing semantic distribution variance. Despite their remarkable achievements, these methods often use coarse semantic representations to model graphs, mainly due to ignoring non-informative elements and failing to focus on precise semantic alignment. Additionally, the generation of coarse graphs inherently introduces abnormal nodes, posing challenges and potentially biasing domain adaptation outcomes. Consequently, we propose a framework, which utilizes the Graph Generation to enhance the quality of DAOD (\method{}). Specifically, we introduce a Node Refinement module that utilizes a memory bank to reconstruct noisy sampled nodes while applying contrastive regularization to noisy features. To enhance semantic alignment, we propose separating domain-specific styles from category invariance encoded within graph covariances, which allows us to selectively remove domain-specific styles while preserving category-invariant information, thus facilitating more accurate semantic alignment across different domains. Furthermore, we propose a Graph Optimization adaptor, leveraging variational inference to mitigate the impact of abnormal nodes. Extensive experimentation across three adaptation benchmarks validates that \method{} achieves state-of-the-art performance in the task of unsupervised domain adaptation.
Authors: Sergio Y. Hayashi, Nina S. T. Hirata
Abstract: Deep neural networks are largely used for complex prediction tasks. There is plenty of empirical evidence of their successful end-to-end training for a diversity of tasks. Success is often measured based solely on the final performance of the trained network, and explanations on when, why and how they work are less emphasized. In this paper we study encoder-decoder recurrent neural networks with attention mechanisms for the task of reading handwritten chess scoresheets. Rather than prediction performance, our concern is to better understand how learning occurs in these type of networks. We characterize the task in terms of three subtasks, namely input-output alignment, sequential pattern recognition, and handwriting recognition, and experimentally investigate which factors affect their learning. We identify competition, collaboration and dependence relations between the subtasks, and argue that such knowledge might help one to better balance factors to properly train a network.
Authors: Sam Sartor, Pieter Peers
Abstract: We formulate SVBRDF estimation from photographs as a diffusion task. To model the distribution of spatially varying materials, we first train a novel unconditional SVBRDF diffusion backbone model on a large set of 312,165 synthetic spatially varying material exemplars. This SVBRDF diffusion backbone model, named MatFusion, can then serve as a basis for refining a conditional diffusion model to estimate the material properties from a photograph under controlled or uncontrolled lighting. Our backbone MatFusion model is trained using only a loss on the reflectance properties, and therefore refinement can be paired with more expensive rendering methods without the need for backpropagation during training. Because the conditional SVBRDF diffusion models are generative, we can synthesize multiple SVBRDF estimates from the same input photograph from which the user can select the one that best matches the users' expectation. We demonstrate the flexibility of our method by refining different SVBRDF diffusion models conditioned on different types of incident lighting, and show that for a single photograph under colocated flash lighting our method achieves equal or better accuracy than existing SVBRDF estimation methods.
Authors: Rishit Dagli, Shivesh Prakash, Robert Wu, Houman Khosravani
Abstract: Generating combined visual and auditory sensory experiences is critical for the consumption of immersive content. Recent advances in neural generative models have enabled the creation of high-resolution content across multiple modalities such as images, text, speech, and videos. Despite these successes, there remains a significant gap in the generation of high-quality spatial audio that complements generated visual content. Furthermore, current audio generation models excel in either generating natural audio or speech or music but fall short in integrating spatial audio cues necessary for immersive experiences. In this work, we introduce SEE-2-SOUND, a zero-shot approach that decomposes the task into (1) identifying visual regions of interest; (2) locating these elements in 3D space; (3) generating mono-audio for each; and (4) integrating them into spatial audio. Using our framework, we demonstrate compelling results for generating spatial audio for high-quality videos, images, and dynamic images from the internet, as well as media generated by learned approaches.
Authors: Zhenyu Li, Shariq Farooq Bhat, Peter Wonka
Abstract: This paper introduces PatchRefiner, an advanced framework for metric single image depth estimation aimed at high-resolution real-domain inputs. While depth estimation is crucial for applications such as autonomous driving, 3D generative modeling, and 3D reconstruction, achieving accurate high-resolution depth in real-world scenarios is challenging due to the constraints of existing architectures and the scarcity of detailed real-world depth data. PatchRefiner adopts a tile-based methodology, reconceptualizing high-resolution depth estimation as a refinement process, which results in notable performance enhancements. Utilizing a pseudo-labeling strategy that leverages synthetic data, PatchRefiner incorporates a Detail and Scale Disentangling (DSD) loss to enhance detail capture while maintaining scale accuracy, thus facilitating the effective transfer of knowledge from synthetic to real-world data. Our extensive evaluations demonstrate PatchRefiner's superior performance, significantly outperforming existing benchmarks on the Unreal4KStereo dataset by 18.1% in terms of the root mean squared error (RMSE) and showing marked improvements in detail accuracy and consistent scale estimation on diverse real-world datasets like CityScape, ScanNet++, and ETH3D.
Authors: Manvik Pasula, Pramit Saha
Abstract: This paper introduces a simple yet effective strategy for exercise classification and muscle group activation prediction (MGAP). These tasks have significant implications for personal fitness, facilitating more affordable, accessible, safer, and simpler exercise routines. This is particularly relevant for novices and individuals with disabilities. Previous research in the field is mostly dominated by the reliance on mounted sensors and a limited scope of exercises, reducing practicality for everyday use. Furthermore, existing MGAP methodologies suffer from a similar dependency on sensors and a restricted range of muscle groups, often excluding strength training exercises, which are pivotal for a comprehensive fitness regimen. Addressing these limitations, our research employs a video-based deep learning framework that encompasses a broad spectrum of exercises and muscle groups, including those vital for strength training. Utilizing the "Workout/Exercises Video" dataset, our approach integrates the X3D and SlowFast video activity recognition models in an effective way to enhance exercise classification and MGAP performance. Our findings demonstrate that this hybrid method obtained via weighted ensemble outperforms existing baseline models in accuracy. Pretrained models play a crucial role in enhancing overall performance, with optimal channel reduction values for the SlowFast model identified near 10. Through an ablation study that explores fine-tuning, we further elucidate the interrelation between the two tasks. Our composite model, a weighted-average ensemble of X3D and SlowFast, sets a new benchmark in both exercise classification and MGAP across all evaluated categories, offering a robust solution to the limitations of previous approaches.
Authors: Ruiyi Zhang, Yanzhe Zhang, Jian Chen, Yufan Zhou, Jiuxiang Gu, Changyou Chen, Tong Sun
Abstract: Large multimodal language models have shown remarkable proficiency in understanding and editing images. However, a majority of these visually-tuned models struggle to comprehend the textual content embedded in images, primarily due to the limitation of training data. In this work, we introduce TRINS: a Text-Rich image INStruction dataset, with the objective of enhancing the reading ability of the multimodal large language model. TRINS is built upon LAION using hybrid data annotation strategies that include machine-assisted and human-assisted annotation processes. It contains 39,153 text-rich images, captions, and 102,437 questions. Specifically, we show that the number of words per annotation in TRINS is significantly longer than that of related datasets, providing new challenges. Furthermore, we introduce a simple and effective architecture, called a Language-vision Reading Assistant (LaRA), which is good at understanding textual content within images. LaRA outperforms existing state-of-the-art multimodal large language models on the TRINS dataset, as well as other classical benchmarks. Lastly, we conducted a comprehensive evaluation with TRINS on various text-rich image understanding and generation tasks, demonstrating its effectiveness.
Authors: Estefania Alfaro-Mejia, Carlos J Delgado, Vidya Manian
Abstract: Spectral Unmixing is an important technique in remote sensing used to analyze hyperspectral images to identify endmembers and estimate abundance maps. Over the past few decades, performance of techniques for endmember extraction and fractional abundance map estimation have significantly improved. This article presents an ensemble model workflow called Autoencoder Graph Ensemble Model (AEGEM) designed to extract endmembers and fractional abundance maps. An elliptical kernel is applied to measure spectral distances, generating the adjacency matrix within the elliptical neighborhood. This information is used to construct an elliptical graph, with centroids as senders and remaining pixels within the geometry as receivers. The next step involves stacking abundance maps, senders, and receivers as inputs to a Graph Convolutional Network, which processes this input to refine abundance maps. Finally, an ensemble decision-making process determines the best abundance maps based on root mean square error metric. The proposed AEGEM is assessed with benchmark datasets such as Samson, Jasper, and Urban, outperforming results obtained by baseline algorithms. For the Samson dataset, AEGEM excels in three abundance maps: water, tree and soil yielding values of 0.081, 0.158, and 0.182, respectively. For the Jasper dataset, results are improved for the tree and water endmembers with values of 0.035 and 0.060 in that order, as well as for the mean average of the spectral angle distance metric 0.109. For the Urban dataset, AEGEM outperforms previous results for the abundance maps of roof and asphalt, achieving values of 0.135 and 0.240, respectively. Additionally, for the endmembers of grass and roof, AEGEM achieves values of 0.063 and 0.094.
Authors: James Lowman, Kelly Liu Zheng, Roydon Fraser, Jesse Van Griensven The, Mojtaba Valipour
Abstract: SeeFar is an evolving collection of multi-resolution satellite images from public and commercial satellites. We specifically curated this dataset for training geospatial foundation models, unconstrained by satellite type. In recent years, advances in technology have made satellite imagery more accessible than ever. More earth-observing satellites have been launched in the last five years than in the previous fifty. Modern commercial satellites now offer up to 100 times the spatial resolution of public access satellites. However, the high cost and limited historical availability of commercial satellite imagery is a barrier to the training of foundational models, impacting what images can be used during inference. The SeeFar dataset represents a step towards training models that are satellite-agnostic by combining multi-resolution commercial and public access pre-processed images. This will enable users to utilize historical data alongside higher-resolution, more expensive satellite imagery, offering greater flexibility during inference. To achieve this, we describe a process for standardizing data from diverse satellite sources, normalizing different data formats, and aligning spectral bands to enhance interoperability. The SeeFar dataset includes images at a resolution of 384x384 pixels, spanning four spectral bands (Blue, Green, Red, and Near-Infrared) and expanding spatial resolutions (starting with 30, 10, 1.5, and 1.0 meters), all in cloud-optimized GeoTIFF format. It also provides consistent and comprehensive metadata to enhance data transparency and reliability. By aggregating data from multiple sources, SeeFar makes processed and consistent satellite data accessible to a wider range of users - from researchers to policymakers - fostering competition and innovation in satellite imagery analysis. The dataset is available at \url{coastalcarbon.ai/seefar}.
Authors: Khiem Le, Zhichun Guo, Kaiwen Dong, Xiaobao Huang, Bozhao Nan, Roshni Iyer, Xiangliang Zhang, Olaf Wiest, Wei Wang, Nitesh V. Chawla
Abstract: Recently, Large Language Models (LLMs) with their strong task-handling capabilities have shown remarkable advancements across a spectrum of fields, moving beyond natural language understanding. However, their proficiency within the chemistry domain remains restricted, especially in solving professional molecule-related tasks. This challenge is attributed to their inherent limitations in comprehending molecules using only common textual representations, i.e., SMILES strings. In this study, we seek to enhance the ability of LLMs to comprehend molecules by designing and equipping them with a multi-modal external module, namely MolX. In particular, instead of directly using a SMILES string to represent a molecule, we utilize specific encoders to extract fine-grained features from both SMILES string and 2D molecular graph representations for feeding into an LLM. Moreover, a human-defined molecular fingerprint is incorporated to leverage its embedded domain knowledge. Then, to establish an alignment between MolX and the LLM's textual input space, the whole model in which the LLM is frozen, is pre-trained with a versatile strategy including a diverse set of tasks. Extensive experimental evaluations demonstrate that our proposed method only introduces a small number of trainable parameters while outperforming baselines on various downstream molecule-related tasks ranging from molecule-to-text translation to retrosynthesis, with and without fine-tuning the LLM.
Authors: Jason Wu, Ziqi Wang, Xiaomin Ouyang, Ho Lyun Jeong, Colin Samplawski, Lance Kaplan, Benjamin Marlin, Mani Srivastava
Abstract: Localization is a critical technology for various applications ranging from navigation and surveillance to assisted living. Localization systems typically fuse information from sensors viewing the scene from different perspectives to estimate the target location while also employing multiple modalities for enhanced robustness and accuracy. Recently, such systems have employed end-to-end deep neural models trained on large datasets due to their superior performance and ability to handle data from diverse sensor modalities. However, such neural models are often trained on data collected from a particular set of sensor poses (i.e., locations and orientations). During real-world deployments, slight deviations from these sensor poses can result in extreme inaccuracies. To address this challenge, we introduce FlexLoc, which employs conditional neural networks to inject node perspective information to adapt the localization pipeline. Specifically, a small subset of model weights are derived from node poses at run time, enabling accurate generalization to unseen perspectives with minimal additional overhead. Our evaluations on a multimodal, multiview indoor tracking dataset showcase that FlexLoc improves the localization accuracy by almost 50% in the zero-shot case (no calibration data available) compared to the baselines. The source code of FlexLoc is available at https://github.com/nesl/FlexLoc.
Authors: Dong Zhao, Shuang Wang, Qi Zang, Licheng Jiao, Nicu Sebe, Zhun Zhong
Abstract: We study source-free unsupervised domain adaptation (SFUDA) for semantic segmentation, which aims to adapt a source-trained model to the target domain without accessing the source data. Many works have been proposed to address this challenging problem, among which uncertainty-based self-training is a predominant approach. However, without comprehensive denoising mechanisms, they still largely fall into biased estimates when dealing with different domains and confirmation bias. In this paper, we observe that pseudo-label noise is mainly contained in unstable samples in which the predictions of most pixels undergo significant variations during self-training. Inspired by this, we propose a novel mechanism to denoise unstable samples with stable ones. Specifically, we introduce the Stable Neighbor Denoising (SND) approach, which effectively discovers highly correlated stable and unstable samples by nearest neighbor retrieval and guides the reliable optimization of unstable samples by bi-level learning. Moreover, we compensate for the stable set by object-level object paste, which can further eliminate the bias caused by less learned classes. Our SND enjoys two advantages. First, SND does not require a specific segmentor structure, endowing its universality. Second, SND simultaneously addresses the issues of class, domain, and confirmation biases during adaptation, ensuring its effectiveness. Extensive experiments show that SND consistently outperforms state-of-the-art methods in various SFUDA semantic segmentation settings. In addition, SND can be easily integrated with other approaches, obtaining further improvements.
Authors: Jan-Martin O. Steitz, Stefan Roth
Abstract: Adapters provide an efficient and lightweight mechanism for adapting trained transformer models to a variety of different tasks. However, they have often been found to be outperformed by other adaptation mechanisms, including low-rank adaptation. In this paper, we provide an in-depth study of adapters, their internal structure, as well as various implementation choices. We uncover pitfalls for using adapters and suggest a concrete, improved adapter architecture, called Adapter+, that not only outperforms previous adapter implementations but surpasses a number of other, more complex adaptation mechanisms in several challenging settings. Despite this, our suggested adapter is highly robust and, unlike previous work, requires little to no manual intervention when addressing a novel scenario. Adapter+ reaches state-of-the-art average accuracy on the VTAB benchmark, even without a per-task hyperparameter optimization.
Authors: Jikai Wang, Qifan Zhang, Yu-Wei Chao, Bowen Wen, Xiaohu Guo, Yu Xiang
Abstract: We introduce a data capture system and a new dataset named HO-Cap that can be used to study 3D reconstruction and pose tracking of hands and objects in videos. The capture system uses multiple RGB-D cameras and a HoloLens headset for data collection, avoiding the use of expensive 3D scanners or mocap systems. We propose a semi-automatic method to obtain annotations of shape and pose of hands and objects in the collected videos, which significantly reduces the required annotation time compared to manual labeling. With this system, we captured a video dataset of humans using objects to perform different tasks, as well as simple pick-and-place and handover of an object from one hand to the other, which can be used as human demonstrations for embodied AI and robot manipulation research. Our data capture setup and annotation framework can be used by the community to reconstruct 3D shapes of objects and human hands and track their poses in videos.
Authors: Haochuan Jiang, Guanyu Yang, Fei Cheng, Kaizhu Huang
Abstract: Synthesizing Chinese characters with consistent style using few stylized examples is challenging. Existing models struggle to generate arbitrary style characters with limited examples. In this paper, we propose the Generalized W-Net, a novel class of W-shaped architectures that addresses this. By incorporating Adaptive Instance Normalization and introducing multi-content, our approach can synthesize Chinese characters in any desired style, even with limited examples. It handles seen and unseen styles during training and can generate new character contents. Experimental results demonstrate the effectiveness of our approach.
Authors: Kiran Kokilepersaud, Yavuz Yarici, Mohit Prabhushankar, Ghassan AlRegib
Abstract: In this work, we propose a novel supervised contrastive loss that enables the integration of taxonomic hierarchy information during the representation learning process. A supervised contrastive loss operates by enforcing that images with the same class label (positive samples) project closer to each other than images with differing class labels (negative samples). The advantage of this approach is that it directly penalizes the structure of the representation space itself. This enables greater flexibility with respect to encoding semantic concepts. However, the standard supervised contrastive loss only enforces semantic structure based on the downstream task (i.e. the class label). In reality, the class label is only one level of a \emph{hierarchy of different semantic relationships known as a taxonomy}. For example, the class label is oftentimes the species of an animal, but between different classes there are higher order relationships such as all animals with wings being ``birds". We show that by explicitly accounting for these relationships with a weighting penalty in the contrastive loss we can out-perform the supervised contrastive loss. Additionally, we demonstrate the adaptability of the notion of a taxonomy by integrating our loss into medical and noise-based settings that show performance improvements by as much as 7%.
Authors: Yuanhao Zhai, Kevin Lin, Zhengyuan Yang, Linjie Li, Jianfeng Wang, Chung-Ching Lin, David Doermann, Junsong Yuan, Lijuan Wang
Abstract: Image diffusion distillation achieves high-fidelity generation with very few sampling steps. However, applying these techniques directly to video diffusion often results in unsatisfactory frame quality due to the limited visual quality in public video datasets. This affects the performance of both teacher and student video diffusion models. Our study aims to improve video diffusion distillation while improving frame appearance using abundant high-quality image data. We propose motion consistency model (MCM), a single-stage video diffusion distillation method that disentangles motion and appearance learning. Specifically, MCM includes a video consistency model that distills motion from the video teacher model, and an image discriminator that enhances frame appearance to match high-quality image data. This combination presents two challenges: (1) conflicting frame learning objectives, as video distillation learns from low-quality video frames while the image discriminator targets high-quality images; and (2) training-inference discrepancies due to the differing quality of video samples used during training and inference. To address these challenges, we introduce disentangled motion distillation and mixed trajectory distillation. The former applies the distillation objective solely to the motion representation, while the latter mitigates training-inference discrepancies by mixing distillation trajectories from both the low- and high-quality video domains. Extensive experiments show that our MCM achieves the state-of-the-art video diffusion distillation performance. Additionally, our method can enhance frame quality in video diffusion models, producing frames with high aesthetic scores or specific styles without corresponding video data.
Authors: Shuaiyi Huang, Saksham Suri, Kamal Gupta, Sai Saketh Rambhatla, Ser-nam Lim, Abhinav Shrivastava
Abstract: Video instance segmentation requires classifying, segmenting, and tracking every object across video frames. Unlike existing approaches that rely on masks, boxes, or category labels, we propose UVIS, a novel Unsupervised Video Instance Segmentation (UVIS) framework that can perform video instance segmentation without any video annotations or dense label-based pretraining. Our key insight comes from leveraging the dense shape prior from the self-supervised vision foundation model DINO and the openset recognition ability from the image-caption supervised vision-language model CLIP. Our UVIS framework consists of three essential steps: frame-level pseudo-label generation, transformer-based VIS model training, and query-based tracking. To improve the quality of VIS predictions in the unsupervised setup, we introduce a dual-memory design. This design includes a semantic memory bank for generating accurate pseudo-labels and a tracking memory bank for maintaining temporal consistency in object tracks. We evaluate our approach on three standard VIS benchmarks, namely YoutubeVIS-2019, YoutubeVIS-2021, and Occluded VIS. Our UVIS achieves 21.1 AP on YoutubeVIS-2019 without any video annotations or dense pretraining, demonstrating the potential of our unsupervised VIS framework.
Authors: Zigeng Chen, Xinyin Ma, Gongfan Fang, Zhenxiong Tan, Xinchao Wang
Abstract: Diffusion models have garnered significant interest from the community for their great generative ability across various applications. However, their typical multi-step sequential-denoising nature gives rise to high cumulative latency, thereby precluding the possibilities of parallel computation. To address this, we introduce AsyncDiff, a universal and plug-and-play acceleration scheme that enables model parallelism across multiple devices. Our approach divides the cumbersome noise prediction model into multiple components, assigning each to a different device. To break the dependency chain between these components, it transforms the conventional sequential denoising into an asynchronous process by exploiting the high similarity between hidden states in consecutive diffusion steps. Consequently, each component is facilitated to compute in parallel on separate devices. The proposed strategy significantly reduces inference latency while minimally impacting the generative quality. Specifically, for the Stable Diffusion v2.1, AsyncDiff achieves a 2.7x speedup with negligible degradation and a 4.0x speedup with only a slight reduction of 0.38 in CLIP Score, on four NVIDIA A5000 GPUs. Our experiments also demonstrate that AsyncDiff can be readily applied to video diffusion models with encouraging performances. The code is available at https://github.com/czg1225/AsyncDiff.
Authors: Yavuz Yarici, Kiran Kokilepersaud, Mohit Prabhushankar, Ghassan AlRegib
Abstract: Self-supervised models create representation spaces that lack clear semantic meaning. This interpretability problem of representations makes traditional explainability methods ineffective in this context. In this paper, we introduce a novel method to analyze representation spaces using three key perceptual components: color, shape, and texture. We employ selective masking of these components to observe changes in representations, resulting in distinct importance maps for each. In scenarios, where labels are absent, these importance maps provide more intuitive explanations as they are integral to the human visual system. Our approach enhances the interpretability of the representation space, offering explanations that resonate with human visual perception. We analyze how different training objectives create distinct representation spaces using perceptual components. Additionally, we examine the representation of images across diverse image domains, providing insights into the role of these components in different contexts.
Authors: Wang Yao, Muhammad Ali Farooq, Joseph Lemley, Peter Corcoran
Abstract: The ability to accurately recognize an individual's face with respect to human aging factor holds significant importance for various private as well as government sectors such as customs and public security bureaus, passport office, and national database systems. Therefore, developing a robust age-invariant face recognition system is of crucial importance to address the challenges posed by ageing and maintain the reliability and accuracy of facial recognition technology. In this research work, the focus is to explore the feasibility of utilizing synthetic ageing data to improve the robustness of face recognition models that can eventually help in recognizing people at broader age intervals. To achieve this, we first design set of experiments to evaluate state-of-the-art synthetic ageing methods. In the next stage we explore the effect of age intervals on a current deep learning-based face recognition algorithm by using synthetic ageing data as well as real ageing data to perform rigorous training and validation. Moreover, these synthetic age data have been used in facilitating face recognition algorithms. Experimental results show that the recognition rate of the model trained on synthetic ageing images is 3.33% higher than the results of the baseline model when tested on images with an age gap of 40 years, which prove the potential of synthetic age data which has been quantified to enhance the performance of age-invariant face recognition systems.
Authors: Zeinab Abboud, Herve Lombaert, Samuel Kadoury
Abstract: Efficiently quantifying predictive uncertainty in medical images remains a challenge. While Bayesian neural networks (BNN) offer predictive uncertainty, they require substantial computational resources to train. Although Bayesian approximations such as ensembles have shown promise, they still suffer from high training and inference costs. Existing approaches mainly address the costs of BNN inference post-training, with little focus on improving training efficiency and reducing parameter complexity. This study introduces a training procedure for a sparse (partial) Bayesian network. Our method selectively assigns a subset of parameters as Bayesian by assessing their deterministic saliency through gradient sensitivity analysis. The resulting network combines deterministic and Bayesian parameters, exploiting the advantages of both representations to achieve high task-specific performance and minimize predictive uncertainty. Demonstrated on multi-label ChestMNIST for classification and ISIC, LIDC-IDRI for segmentation, our approach achieves competitive performance and predictive uncertainty estimation by reducing Bayesian parameters by over 95\%, significantly reducing computational expenses compared to fully Bayesian and ensemble methods.
Authors: Shangjie Xue, Jesse Dill, Pranay Mathur, Frank Dellaert, Panagiotis Tsiotra, Danfei Xu
Abstract: This paper presents Neural Visibility Field (NVF), a novel uncertainty quantification method for Neural Radiance Fields (NeRF) applied to active mapping. Our key insight is that regions not visible in the training views lead to inherently unreliable color predictions by NeRF at this region, resulting in increased uncertainty in the synthesized views. To address this, we propose to use Bayesian Networks to composite position-based field uncertainty into ray-based uncertainty in camera observations. Consequently, NVF naturally assigns higher uncertainty to unobserved regions, aiding robots to select the most informative next viewpoints. Extensive evaluations show that NVF excels not only in uncertainty quantification but also in scene reconstruction for active mapping, outperforming existing methods.
Authors: Weiwei Duan, Luping Ji, Shengjia Chen, Sicheng Zhu, Mao Ye
Abstract: Moving infrared small target detection presents significant challenges due to tiny target sizes and low contrast against backgrounds. Currently-existing methods primarily focus on extracting target features only from the spatial-temporal domain. For further enhancing feature representation, more information domains such as frequency are believed to be potentially valuable. To extend target feature learning, we propose a new Triple-domain Strategy (Tridos) with the frequency-aware memory enhancement on the spatial-temporal domain. In our scheme, it effectively detaches and enhances frequency features by a local-global frequency-aware module with Fourier transform. Inspired by the human visual system, our memory enhancement aims to capture the target spatial relations between video frames. Furthermore, it encodes temporal dynamics motion features via differential learning and residual enhancing. Additionally, we further design a residual compensation unit to reconcile possible cross-domain feature mismatches. To our best knowledge, our Tridos is the first work to explore target feature learning comprehensively in spatial-temporal-frequency domains. The extensive experiments on three datasets (DAUB, ITSDT-15K, and IRDST) validate that our triple-domain learning scheme could be obviously superior to state-of-the-art ones. Source codes are available at https://github.com/UESTC-nnLab/Tridos.
Authors: Weiqing Xiao, Wei Zhao
Abstract: Due to the difficulty in obtaining real samples and ground truth, the generalization performance and the fine-tuned performance are critical for the feasibility of stereo matching methods in real-world applications. However, the presence of substantial disparity distributions and density variations across different datasets presents significant challenges for the generalization and fine-tuning of the model. In this paper, we propose a novel stereo matching method, called SR-Stereo, which mitigates the distributional differences across different datasets by predicting the disparity clips and uses a loss weight related to the regression target scale to improve the accuracy of the disparity clips. Moreover, this stepwise regression architecture can be easily extended to existing iteration-based methods to improve the performance without changing the structure. In addition, to mitigate the edge blurring of the fine-tuned model on sparse ground truth, we propose Domain Adaptation Based on Pre-trained Edges (DAPE). Specifically, we use the predicted disparity and RGB image to estimate the edge map of the target domain image. The edge map is filtered to generate edge map background pseudo-labels, which together with the sparse ground truth disparity on the target domain are used as a supervision to jointly fine-tune the pre-trained stereo matching model. These proposed methods are extensively evaluated on SceneFlow, KITTI, Middbury 2014 and ETH3D. The SR-Stereo achieves competitive disparity estimation performance and state-of-the-art cross-domain generalisation performance. Meanwhile, the proposed DAPE significantly improves the disparity estimation performance of fine-tuned models, especially in the textureless and detail regions.
Authors: Ping Liu, Qiqi Tao, Joey Tianyi Zhou
Abstract: This survey addresses the critical challenge of deepfake detection amidst the rapid advancements in artificial intelligence. As AI-generated media, including video, audio and text, become more realistic, the risk of misuse to spread misinformation and commit identity fraud increases. Focused on face-centric deepfakes, this work traces the evolution from traditional single-modality methods to sophisticated multi-modal approaches that handle audio-visual and text-visual scenarios. We provide comprehensive taxonomies of detection techniques, discuss the evolution of generative methods from auto-encoders and GANs to diffusion models, and categorize these technologies by their unique attributes. To our knowledge, this is the first survey of its kind. We also explore the challenges of adapting detection methods to new generative models and enhancing the reliability and robustness of deepfake detectors, proposing directions for future research. This survey offers a detailed roadmap for researchers, supporting the development of technologies to counter the deceptive use of AI in media creation, particularly facial forgery. A curated list of all related papers can be found at \href{https://github.com/qiqitao77/Comprehensive-Advances-in-Deepfake-Detection-Spanning-Diverse-Modalities}{https://github.com/qiqitao77/Awesome-Comprehensive-Deepfake-Detection}.
URLs: https://github.com/qiqitao77/Comprehensive-Advances-in-Deepfake-Detection-Spanning-Diverse-Modalities, https://github.com/qiqitao77/Awesome-Comprehensive-Deepfake-Detection
Authors: Kailas Dayanandan, Anand Sinha, Brejesh Lall
Abstract: The dual thinking framework considers fast, intuitive processing and slower, logical processing. The perception of dual thinking in vision requires images where inferences from intuitive and logical processing differ. We introduce an adversarial dataset to provide evidence for the dual thinking framework in human vision, which also aids in studying the qualitative behavior of deep learning models. Our study also addresses a major criticism of using classification models as computational models of human vision by using instance segmentation models that localize objects. The evidence underscores the importance of shape in identifying instances in human vision and shows that deep learning models lack an understanding of sub-structures, as indicated by errors related to the position and number of sub-components. Additionally, the similarity in errors made by models and intuitive human processing indicates that models only address intuitive thinking in human vision.
Authors: Xin Yuan, Rana Hanocka, Michael Maire
Abstract: We cast multiview reconstruction from unknown pose as a generative modeling problem. From a collection of unannotated 2D images of a scene, our approach simultaneously learns both a network to predict camera pose from 2D image input, as well as the parameters of a Neural Radiance Field (NeRF) for the 3D scene. To drive learning, we wrap both the pose prediction network and NeRF inside a Denoising Diffusion Probabilistic Model (DDPM) and train the system via the standard denoising objective. Our framework requires the system accomplish the task of denoising an input 2D image by predicting its pose and rendering the NeRF from that pose. Learning to denoise thus forces the system to concurrently learn the underlying 3D NeRF representation and a mapping from images to camera extrinsic parameters. To facilitate the latter, we design a custom network architecture to represent pose as a distribution, granting implicit capacity for discovering view correspondences when trained end-to-end for denoising alone. This technique allows our system to successfully build NeRFs, without pose knowledge, for challenging scenes where competing methods fail. At the conclusion of training, our learned NeRF can be extracted and used as a 3D scene model; our full system can be used to sample novel camera poses and generate novel-view images.
Authors: Tiancheng Gu, Kaicheng Yang, Xiang An, Ziyong Feng, Dongnan Liu, Weidong Cai, Jiankang Deng
Abstract: Contrastive Language-Image Pre-training (CLIP) has significantly improved performance in various vision-language tasks by expanding the dataset with image-text pairs obtained from websites. This paper further explores CLIP from the perspectives of data and model architecture. To address the prevalence of noisy data and enhance the quality of large-scale image-text data crawled from the internet, we introduce a diverse description generation framework that can leverage Large Language Models (LLMs) to synthesize and refine content from web-based texts, synthetic captions, and detection tags. Furthermore, we propose RWKV-CLIP, the first RWKV-driven vision-language representation learning model that combines the effective parallel training of transformers with the efficient inference of RNNs. Comprehensive experiments across various model scales and pre-training datasets demonstrate that RWKV-CLIP is a robust and efficient vision-language representation learner, it achieves state-of-the-art performance in several downstream tasks, including linear probe, zero-shot classification, and zero-shot image-text retrieval. To facilitate future research, the code and pre-trained models are released at https://github.com/deepglint/RWKV-CLIP
Authors: Zhenxin Li, Kailin Li, Shihao Wang, Shiyi Lan, Zhiding Yu, Yishen Ji, Zhiqi Li, Ziyue Zhu, Jan Kautz, Zuxuan Wu, Yu-Gang Jiang, Jose M. Alvarez
Abstract: We propose Hydra-MDP, a novel paradigm employing multiple teachers in a teacher-student model. This approach uses knowledge distillation from both human and rule-based teachers to train the student model, which features a multi-head decoder to learn diverse trajectory candidates tailored to various evaluation metrics. With the knowledge of rule-based teachers, Hydra-MDP learns how the environment influences the planning in an end-to-end manner instead of resorting to non-differentiable post-processing. This method achieves the $1^{st}$ place in the Navsim challenge, demonstrating significant improvements in generalization across diverse driving environments and conditions. Code will be available at \url{https://github.com/woxihuanjiangguo/Hydra-MDP}
Authors: Junfei Yi, Jianxu Mao, Tengfei Liu, Mingjie Li, Hanyu Gu, Hui Zhang, Xiaojun Chang, Yaonan Wang
Abstract: Knowledge distillation (KD) is a widely adopted and effective method for compressing models in object detection tasks. Particularly, feature-based distillation methods have shown remarkable performance. Existing approaches often ignore the uncertainty in the teacher model's knowledge, which stems from data noise and imperfect training. This limits the student model's ability to learn latent knowledge, as it may overly rely on the teacher's imperfect guidance. In this paper, we propose a novel feature-based distillation paradigm with knowledge uncertainty for object detection, termed "Uncertainty Estimation-Discriminative Knowledge Extraction-Knowledge Transfer (UET)", which can seamlessly integrate with existing distillation methods. By leveraging the Monte Carlo dropout technique, we introduce knowledge uncertainty into the training process of the student model, facilitating deeper exploration of latent knowledge. Our method performs effectively during the KD process without requiring intricate structures or extensive computational resources. Extensive experiments validate the effectiveness of our proposed approach across various distillation strategies, detectors, and backbone architectures. Specifically, following our proposed paradigm, the existing FGD method achieves state-of-the-art (SoTA) performance, with ResNet50-based GFL achieving 44.1% mAP on the COCO dataset, surpassing the baselines by 3.9%.
Authors: Xin Jin, Chunle Guo, Xiaoming Li, Zongsheng Yue, Chongyi Li, Shangchen Zhou, Ruicheng Feng, Yuekun Dai, Peiqing Yang, Chen Change Loy, Ruoqi Li, Chang Liu, Ziyi Wang, Yao Du, Jingjing Yang, Long Bao, Heng Sun, Xiangyu Kong, Xiaoxia Xing, Jinlong Wu, Yuanyang Xue, Hyunhee Park, Sejun Song, Changho Kim, Jingfan Tan, Wenhan Luo, Zikun Liu, Mingde Qiao, Junjun Jiang, Kui Jiang, Yao Xiao, Chuyang Sun, Jinhui Hu, Weijian Ruan, Yubo Dong, Kai Chen, Hyejeong Jo, Jiahao Qin, Bingjie Han, Pinle Qin, Rui Chai, Pengyuan Wang
Abstract: The increasing demand for computational photography and imaging on mobile platforms has led to the widespread development and integration of advanced image sensors with novel algorithms in camera systems. However, the scarcity of high-quality data for research and the rare opportunity for in-depth exchange of views from industry and academia constrain the development of mobile intelligent photography and imaging (MIPI). Building on the achievements of the previous MIPI Workshops held at ECCV 2022 and CVPR 2023, we introduce our third MIPI challenge including three tracks focusing on novel image sensors and imaging algorithms. In this paper, we summarize and review the Few-shot RAW Image Denoising track on MIPI 2024. In total, 165 participants were successfully registered, and 7 teams submitted results in the final testing phase. The developed solutions in this challenge achieved state-of-the-art erformance on Few-shot RAW Image Denoising. More details of this challenge and the link to the dataset can be found at https://mipichallenge.org/MIPI2024.
Authors: Sooyeon Go, Kyungmook Choi, Minjung Shin, Youngjung Uh
Abstract: As pretrained text-to-image diffusion models have become a useful tool for image synthesis, people want to specify the results in various ways. In this paper, we introduce a method to produce results with the same structure of a target image but painted with colors from a reference image, i.e., appearance transfer, especially following the semantic correspondence between the result and the reference. E.g., the result wing takes color from the reference wing, not the reference head. Existing methods rely on the query-key similarity within self-attention layer, usually producing defective results. To this end, we propose to find semantic correspondences and explicitly rearrange the features according to the semantic correspondences. Extensive experiments show the superiority of our method in various aspects: preserving the structure of the target and reflecting the color from the reference according to the semantic correspondences, even when the two images are not aligned.
Authors: Jiahua Xu, Si Zuo, Chenfeng Wei, Wei Zhou
Abstract: With the rapid proliferation of autonomous driving, there has been a heightened focus on the research of lidar-based 3D semantic segmentation and object detection methodologies, aiming to ensure the safety of traffic participants. In recent decades, learning-based approaches have emerged, demonstrating remarkable performance gains in comparison to conventional algorithms. However, the segmentation and detection tasks have traditionally been examined in isolation to achieve the best precision. To this end, we propose an efficient multi-task learning framework named LiSD which can address both segmentation and detection tasks, aiming to optimize the overall performance. Our proposed LiSD is a voxel-based encoder-decoder framework that contains a hierarchical feature collaboration module and a holistic information aggregation module. Different integration methods are adopted to keep sparsity in segmentation while densifying features for query initialization in detection. Besides, cross-task information is utilized in an instance-aware refinement module to obtain more accurate predictions. Experimental results on the nuScenes dataset and Waymo Open Dataset demonstrate the effectiveness of our proposed model. It is worth noting that LiSD achieves the state-of-the-art performance of 83.3% mIoU on the nuScenes segmentation benchmark for lidar-only methods.
Authors: Zhechao Wang, Peirui Cheng, Pengju Tian, Yuchao Wang, Mingxin Chen, Shujing Duan, Zhirui Wang, Xinming Li, Xian Sun
Abstract: Remote sensing lightweight foundation models have achieved notable success in online perception within remote sensing. However, their capabilities are restricted to performing online inference solely based on their own observations and models, thus lacking a comprehensive understanding of large-scale remote sensing scenarios. To overcome this limitation, we propose a Remote Sensing Distributed Foundation Model (RS-DFM) based on generalized information mapping and interaction. This model can realize online collaborative perception across multiple platforms and various downstream tasks by mapping observations into a unified space and implementing a task-agnostic information interaction strategy. Specifically, we leverage the ground-based geometric prior of remote sensing oblique observations to transform the feature mapping from absolute depth estimation to relative depth estimation, thereby enhancing the model's ability to extract generalized features across diverse heights and perspectives. Additionally, we present a dual-branch information compression module to decouple high-frequency and low-frequency feature information, achieving feature-level compression while preserving essential task-agnostic details. In support of our research, we create a multi-task simulation dataset named AirCo-MultiTasks for multi-UAV collaborative observation. We also conduct extensive experiments, including 3D object detection, instance segmentation, and trajectory prediction. The numerous results demonstrate that our RS-DFM achieves state-of-the-art performance across various downstream tasks.
Authors: Yining Shi, Jiusi Li, Kun Jiang, Ke Wang, Yunlong Wang, Mengmeng Yang, Diange Yang
Abstract: Vision-centric occupancy networks, which represent the surrounding environment with uniform voxels with semantics, have become a new trend for safe driving of camera-only autonomous driving perception systems, as they are able to detect obstacles regardless of their shape and occlusion. Modern occupancy networks mainly focus on reconstructing visible voxels from object surfaces with voxel-wise semantic prediction. Usually, they suffer from inconsistent predictions of one object and mixed predictions for adjacent objects. These confusions may harm the safety of downstream planning modules. To this end, we investigate panoptic segmentation on 3D voxel scenarios and propose an instance-aware occupancy network, PanoSSC. We predict foreground objects and backgrounds separately and merge both in post-processing. For foreground instance grouping, we propose a novel 3D instance mask decoder that can efficiently extract individual objects. we unify geometric reconstruction, 3D semantic segmentation, and 3D instance segmentation into PanoSSC framework and propose new metrics for evaluating panoptic voxels. Extensive experiments show that our method achieves competitive results on SemanticKITTI semantic scene completion benchmark.
Authors: Yining Shi, Kun Jiang, Ke Wang, Kangan Qian, Yunlong Wang, Jiusi Li, Tuopu Wen, Mengmeng Yang, Yiliang Xu, Diange Yang
Abstract: 3D occupancy prediction (Occ) is a rapidly rising challenging perception task in the field of autonomous driving which represents the driving scene as uniformly partitioned 3D voxel grids with semantics. Compared to 3D object detection, grid perception has great advantage of better recognizing irregularly shaped, unknown category, or partially occluded general objects. However, existing 3D occupancy networks (occnets) are both computationally heavy and label-hungry. In terms of model complexity, occnets are commonly composed of heavy Conv3D modules or transformers on the voxel level. In terms of label annotations requirements, occnets are supervised with large-scale expensive dense voxel labels. Model and data inefficiency, caused by excessive network parameters and label annotations requirement, severely hinder the onboard deployment of occnets. This paper proposes an efficient 3d occupancy network (EFFOcc), that targets the minimal network complexity and label requirement while achieving state-of-the-art accuracy. EFFOcc only uses simple 2D operators, and improves Occ accuracy to the state-of-the-art on multiple large-scale benchmarks: Occ3D-nuScenes, Occ3D-Waymo, and OpenOccupancy-nuScenes. On Occ3D-nuScenes benchmark, EFFOcc has only 18.4M parameters, and achieves 50.46 in terms of mean IoU (mIoU), to our knowledge, it is the occnet with minimal parameters compared with related occnets. Moreover, we propose a two-stage active learning strategy to reduce the requirements of labelled data. Active EFFOcc trained with 6\% labelled voxels achieves 47.19 mIoU, which is 95.7% fully supervised performance. The proposed EFFOcc also supports improved vision-only occupancy prediction with the aid of region-decomposed distillation. Code and demo videos will be available at https://github.com/synsin0/EFFOcc.
Authors: Mingqi Gao, Jingnan Luo, Jinyu Yang, Jungong Han, Feng Zheng
Abstract: Motion Expression guided Video Segmentation (MeViS), as an emerging task, poses many new challenges to the field of referring video object segmentation (RVOS). In this technical report, we investigated and validated the effectiveness of static-dominant data and frame sampling on this challenging setting. Our solution achieves a J&F score of 0.5447 in the competition phase and ranks 1st in the MeViS track of the PVUW Challenge. The code is available at: https://github.com/Tapall-AI/MeViS_Track_Solution_2024.
URLs: https://github.com/Tapall-AI/MeViS_Track_Solution_2024.
Authors: Jiamu Sheng, Jingyi Zhou, Jiong Wang, Peng Ye, Jiayuan Fan
Abstract: The effectiveness and efficiency of modeling complex spectral-spatial relations are both crucial for Hyperspectral image (HSI) classification. Most existing methods based on CNNs and transformers still suffer from heavy computational burdens and have room for improvement in capturing the global-local spectral-spatial feature representation. To this end, we propose a novel lightweight parallel design called lightweight dual-stream Mamba-convolution network (DualMamba) for HSI classification. Specifically, a parallel lightweight Mamba and CNN block are first developed to extract global and local spectral-spatial features. First, the cross-attention spectral-spatial Mamba module is proposed to leverage the global modeling of Mamba at linear complexity. Within this module, dynamic positional embedding is designed to enhance the spatial location information of visual sequences. The lightweight spectral/spatial Mamba blocks comprise an efficient scanning strategy and a lightweight Mamba design to efficiently extract global spectral-spatial features. And the cross-attention spectral-spatial fusion is designed to learn cross-correlation and fuse spectral-spatial features. Second, the lightweight spectral-spatial residual convolution module is proposed with lightweight spectral and spatial branches to extract local spectral-spatial features through residual learning. Finally, the adaptive global-local fusion is proposed to dynamically combine global Mamba features and local convolution features for a global-local spectral-spatial representation. Compared with state-of-the-art HSI classification methods, experimental results demonstrate that DualMamba achieves significant classification accuracy on three public HSI datasets and a superior reduction in model parameters and floating point operations (FLOPs).
Authors: Huahui Yi, Xiaofei Wang, Kang Li, Chao Li
Abstract: Multimodal learning, integrating histology images and genomics, promises to enhance precision oncology with comprehensive views at microscopic and molecular levels. However, existing methods may not sufficiently model the shared or complementary information for more effective integration. In this study, we introduce a Unified Modeling Enhanced Multimodal Learning (UMEML) framework that employs a hierarchical attention structure to effectively leverage shared and complementary features of both modalities of histology and genomics. Specifically, to mitigate unimodal bias from modality imbalance, we utilize a query-based cross-attention mechanism for prototype clustering in the pathology encoder. Our prototype assignment and modularity strategy are designed to align shared features and minimizes modality gaps. An additional registration mechanism with learnable tokens is introduced to enhance cross-modal feature integration and robustness in multimodal unified modeling. Our experiments demonstrate that our method surpasses previous state-of-the-art approaches in glioma diagnosis and prognosis tasks, underscoring its superiority in precision neuro-Oncology.
Authors: Zhongzhen Huang, Yankai Jiang, Rongzhao Zhang, Shaoting Zhang, Xiaofan Zhang
Abstract: Existing promptable segmentation methods in the medical imaging field primarily consider either textual or visual prompts to segment relevant objects, yet they often fall short when addressing anomalies in medical images, like tumors, which may vary greatly in shape, size, and appearance. Recognizing the complexity of medical scenarios and the limitations of textual or visual prompts, we propose a novel dual-prompt schema that leverages the complementary strengths of visual and textual prompts for segmenting various organs and tumors. Specifically, we introduce CAT, an innovative model that Coordinates Anatomical prompts derived from 3D cropped images with Textual prompts enriched by medical domain knowledge. The model architecture adopts a general query-based design, where prompt queries facilitate segmentation queries for mask prediction. To synergize two types of prompts within a unified framework, we implement a ShareRefiner, which refines both segmentation and prompt queries while disentangling the two types of prompts. Trained on a consortium of 10 public CT datasets, CAT demonstrates superior performance in multiple segmentation tasks. Further validation on a specialized in-house dataset reveals the remarkable capacity of segmenting tumors across multiple cancer stages. This approach confirms that coordinating multimodal prompts is a promising avenue for addressing complex scenarios in the medical domain.
Authors: Wenjia Xu, Zijian Yu, Yixu Wang, Jiuniu Wang, Mugen Peng
Abstract: An increasing number of models have achieved great performance in remote sensing tasks with the recent development of Large Language Models (LLMs) and Visual Language Models (VLMs). However, these models are constrained to basic vision and language instruction-tuning tasks, facing challenges in complex remote sensing applications. Additionally, these models lack specialized expertise in professional domains. To address these limitations, we propose a LLM-driven remote sensing intelligent agent named RS-Agent. Firstly, RS-Agent is powered by a large language model (LLM) that acts as its "Central Controller," enabling it to understand and respond to various problems intelligently. Secondly, our RS-Agent integrates many high-performance remote sensing image processing tools, facilitating multi-tool and multi-turn conversations. Thirdly, our RS-Agent can answer professional questions by leveraging robust knowledge documents. We conducted experiments using several datasets, e.g., RSSDIVCS, RSVQA, and DOTAv1. The experimental results demonstrate that our RS-Agent delivers outstanding performance in many tasks, i.e., scene classification, visual question answering, and object counting tasks.
Authors: Xing Zhang, Jiaxi Gu, Haoyu Zhao, Shicong Wang, Hang Xu, Renjing Pei, Songcen Xu, Zuxuan Wu, Yu-Gang Jiang
Abstract: Temporal Video Grounding (TVG) aims to localize a moment from an untrimmed video given the language description. Since the annotation of TVG is labor-intensive, TVG under limited supervision has accepted attention in recent years. The great success of vision-language pre-training guides TVG to follow the traditional "pre-training + fine-tuning" paradigm, however, the pre-training process would suffer from a lack of temporal modeling and fine-grained alignment due to the difference of data nature between pre-train and test. Besides, the large gap between pretext and downstream tasks makes zero-shot testing impossible for the pre-trained model. To avoid the drawbacks of the traditional paradigm, we propose AutoTVG, a new vision-language pre-training paradigm for TVG that enables the model to learn semantic alignment and boundary regression from automatically annotated untrimmed videos. To be specific, AutoTVG consists of a novel Captioned Moment Generation (CMG) module to generate captioned moments from untrimmed videos, and TVGNet with a regression head to predict localization results. Experimental results on Charades-STA and ActivityNet Captions show that, regarding zero-shot temporal video grounding, AutoTVG achieves highly competitive performance with in-distribution methods under out-of-distribution testing, and is superior to existing pre-training frameworks with much less training data.
Authors: Yufei Han, Heng Guo, Koki Fukai, Hiroaki Santo, Boxin Shi, Fumio Okura, Zhanyu Ma, Yunpeng Jia
Abstract: We present NeRSP, a Neural 3D reconstruction technique for Reflective surfaces with Sparse Polarized images. Reflective surface reconstruction is extremely challenging as specular reflections are view-dependent and thus violate the multiview consistency for multiview stereo. On the other hand, sparse image inputs, as a practical capture setting, commonly cause incomplete or distorted results due to the lack of correspondence matching. This paper jointly handles the challenges from sparse inputs and reflective surfaces by leveraging polarized images. We derive photometric and geometric cues from the polarimetric image formation model and multiview azimuth consistency, which jointly optimize the surface geometry modeled via implicit neural representation. Based on the experiments on our synthetic and real datasets, we achieve the state-of-the-art surface reconstruction results with only 6 views as input.
Authors: Sergey Linok, Tatiana Zemskova, Svetlana Ladanova, Roman Titkov, Dmitry Yudin
Abstract: Locating objects referred to in natural language poses a significant challenge for autonomous agents. Existing CLIP-based open-vocabulary methods successfully perform 3D object retrieval with simple (bare) queries but cannot cope with ambiguous descriptions that demand an understanding of object relations. To tackle this problem, we propose a modular approach called BBQ (Beyond Bare Queries), which constructs 3D scene spatial graph representation with metric edges and utilizes a large language model as a human-to-agent interface through our deductive scene reasoning algorithm. BBQ employs robust DINO-powered associations to form 3D objects, an advanced raycasting algorithm to project them to 2D, and a vision-language model to describe them as graph nodes. On Replica and ScanNet datasets, we show that the designed method accurately constructs 3D object-centric maps. We have demonstrated that their quality takes a leading place for open-vocabulary 3D semantic segmentation against other zero-shot methods. Also, we show that leveraging spatial relations is especially effective for scenes containing multiple entities of the same semantic class. On Sr3D and Nr3D benchmarks, our deductive approach demonstrates a significant improvement, enabling retrieving objects by complex queries compared to other state-of-the-art methods. Considering our design solutions, we achieved a processing speed approximately x3 times faster than the closest analog. This promising performance enables our approach for usage in applied intelligent robotics projects. We make the code publicly available at linukc.github.io/bbq/.
Authors: Aoxiong Yin, Haoyuan Li, Kai Shen, Siliang Tang, Yueting Zhuang
Abstract: In this work, we propose a two-stage sign language production (SLP) paradigm that first encodes sign language sequences into discrete codes and then autoregressively generates sign language from text based on the learned codebook. However, existing vector quantization (VQ) methods are fixed-length encodings, overlooking the uneven information density in sign language, which leads to under-encoding of important regions and over-encoding of unimportant regions. To address this issue, we propose a novel dynamic vector quantization (DVA-VAE) model that can dynamically adjust the encoding length based on the information density in sign language to achieve accurate and compact encoding. Then, a GPT-like model learns to generate code sequences and their corresponding durations from spoken language text. Extensive experiments conducted on the PHOENIX14T dataset demonstrate the effectiveness of our proposed method. To promote sign language research, we propose a new large German sign language dataset, PHOENIX-News, which contains 486 hours of sign language videos, audio, and transcription texts.Experimental analysis on PHOENIX-News shows that the performance of our model can be further improved by increasing the size of the training data. Our project homepage is https://t2sgpt-demo.yinaoxiong.cn.
Authors: Che Liu, Zhongwei Wan, Yuqi Wang, Hui Shen, Haozhe Wang, Kangyu Zheng, Mi Zhang, Rossella Arcucci
Abstract: Automatic radiology report generation can significantly benefit the labor-intensive process of report writing by radiologists, especially for 3D radiographs like CT scans, which are crucial for broad clinical diagnostics yet underexplored compared to 2D radiographs. Existing methods often handle 3D volumes either slice-wise or with aggressive downsampling due to current GPU memory limitations, which results in a loss of the inherent 3D nature and critical details. To overcome these issues, we introduce a novel framework that efficiently and effectively generates radiology reports for high-resolution (HR) 3D volumes, based on large language models (LLMs). Specifically, our framework utilizes low-resolution (LR) visual tokens as queries to mine information from HR tokens, preserving detailed HR information while reducing computational costs by only processing HR informed LR visual queries. Further benefiting the field, we curate and release BIMCV-RG, a new dataset with 5,328 HR 3D volumes and paired reports, establishing the first benchmarks for report generation from 3D HR medical images. Our method consistently surpasses existing methods on this benchmark across three different settings: normal-resolution, high-resolution inputs, and zero-shot domain transfer, all at an acceptable computational cost, trainable on a single A100-80G.
Authors: Haoran Wang, Mohit Mendiratta, Christian Theobalt, Adam Kortylewski
Abstract: We introduce FaceGPT, a self-supervised learning framework for Large Vision-Language Models (VLMs) to reason about 3D human faces from images and text. Typical 3D face reconstruction methods are specialized algorithms that lack semantic reasoning capabilities. FaceGPT overcomes this limitation by embedding the parameters of a 3D morphable face model (3DMM) into the token space of a VLM, enabling the generation of 3D faces from both textual and visual inputs. FaceGPT is trained in a self-supervised manner as a model-based autoencoder from in-the-wild images. In particular, the hidden state of LLM is projected into 3DMM parameters and subsequently rendered as 2D face image to guide the self-supervised learning process via image-based reconstruction. Without relying on expensive 3D annotations of human faces, FaceGPT obtains a detailed understanding about 3D human faces, while preserving the capacity to understand general user instructions. Our experiments demonstrate that FaceGPT not only achieves high-quality 3D face reconstructions but also retains the ability for general-purpose visual instruction following. Furthermore, FaceGPT learns fully self-supervised to generate 3D faces based on complex textual inputs, which opens a new direction in human face analysis.
Authors: Mirgahney Mohamed, Harry Jake Cunningham, Marc P. Deisenroth, Lourdes Agapito
Abstract: Human motion generation has paramount importance in computer animation. It is a challenging generative temporal modelling task due to the vast possibilities of human motion, high human sensitivity to motion coherence and the difficulty of accurately generating fine-grained motions. Recently, diffusion methods have been proposed for human motion generation due to their high sample quality and expressiveness. However, generated sequences still suffer from motion incoherence, and are limited to short duration, and simpler motion and take considerable time during inference. To address these limitations, we propose \textit{RecMoDiffuse: Recurrent Flow Diffusion}, a new recurrent diffusion formulation for temporal modelling. Unlike previous work, which applies diffusion to the whole sequence without any temporal dependency, an approach that inherently makes temporal consistency hard to achieve. Our method explicitly enforces temporal constraints with the means of normalizing flow models in the diffusion process and thereby extends diffusion to the temporal dimension. We demonstrate the effectiveness of RecMoDiffuse in the temporal modelling of human motion. Our experiments show that RecMoDiffuse achieves comparable results with state-of-the-art methods while generating coherent motion sequences and reducing the computational overhead in the inference stage.
Authors: Sidun Liu, Peng Qiao, Zongxin Ye, Wenyu Li, Yong Dou
Abstract: Neural Surface Reconstruction learns a Signed Distance Field~(SDF) to reconstruct the 3D model from multi-view images. Previous works adopt voxel-based explicit representation to improve efficiency. However, they ignored the gradient instability of interpolation in the voxel grid, leading to degradation on convergence and smoothness. Besides, previous works entangled the optimization of geometry and radiance, which leads to the deformation of geometry to explain radiance, causing artifacts when reconstructing textured planes. In this work, we reveal that the instability of gradient comes from its discontinuity during trilinear interpolation, and propose to use the interpolated gradient instead of the original analytical gradient to eliminate the discontinuity. Based on gradient interpolation, we propose VoxNeuS, a lightweight surface reconstruction method for computational and memory efficient neural surface reconstruction. Thanks to the explicit representation, the gradient of regularization terms, i.e. Eikonal and curvature loss, are directly solved, avoiding computation and memory-access overhead. Further, VoxNeuS adopts a geometry-radiance disentangled architecture to handle the geometry deformation from radiance optimization. The experimental results show that VoxNeuS achieves better reconstruction quality than previous works. The entire training process takes 15 minutes and less than 3 GB of memory on a single 2080ti GPU.
Authors: Yuqi Cheng, Yunkang Cao, Rui Chen, Weiming Shen
Abstract: Robustness against noisy imaging is crucial for practical image anomaly detection systems. This study introduces a Robust Anomaly Detection (RAD) dataset with free views, uneven illuminations, and blurry collections to systematically evaluate the robustness of current anomaly detection methods. Specifically, RAD aims to identify foreign objects on working platforms as anomalies. The collection process incorporates various sources of imaging noise, such as viewpoint changes, uneven illuminations, and blurry collections, to replicate real-world inspection scenarios. Subsequently, we assess and analyze 11 state-of-the-art unsupervised and zero-shot methods on RAD. Our findings indicate that: 1) Variations in viewpoint, illumination, and blurring affect anomaly detection methods to varying degrees; 2) Methods relying on memory banks and assisted by synthetic anomalies demonstrate stronger robustness; 3) Effectively leveraging the general knowledge of foundational models is a promising avenue for enhancing the robustness of anomaly detection methods.
Authors: Yunfeng Li, Bo Wang, Jiuran Sun, Xueyi Wu, Ye Li
Abstract: Vision camera and sonar are naturally complementary in the underwater environment. Combining the information from two modalities will promote better observation of underwater targets. However, this problem has not received sufficient attention in previous research. Therefore, this paper introduces a new challenging RGB-Sonar (RGB-S) tracking task and investigates how to achieve efficient tracking of an underwater target through the interaction of RGB and sonar modalities. Specifically, we first propose an RGBS50 benchmark dataset containing 50 sequences and more than 87000 high-quality annotated bounding boxes. Experimental results show that the RGBS50 benchmark poses a challenge to currently popular SOT trackers. Second, we propose an RGB-S tracker called SCANet, which includes a spatial cross-attention module (SCAM) consisting of a novel spatial cross-attention layer and two independent global integration modules. The spatial cross-attention is used to overcome the problem of spatial misalignment of between RGB and sonar images. Third, we propose a SOT data-based RGB-S simulation training method (SRST) to overcome the lack of RGB-S training datasets. It converts RGB images into sonar-like saliency images to construct pseudo-data pairs, enabling the model to learn the semantic structure of RGB-S-like data. Comprehensive experiments show that the proposed spatial cross-attention effectively achieves the interaction between RGB and sonar modalities and SCANet achieves state-of-the-art performance on the proposed benchmark. The code is available at https://github.com/LiYunfengLYF/RGBS50.
Authors: Ioanna Ntinou, Enrique Sanchez, Georgios Tzimiropoulos
Abstract: This paper is on long-term video understanding where the goal is to recognise human actions over long temporal windows (up to minutes long). In prior work, long temporal context is captured by constructing a long-term memory bank consisting of past and future video features which are then integrated into standard (short-term) video recognition backbones through the use of attention mechanisms. Two well-known problems related to this approach are the quadratic complexity of the attention operation and the fact that the whole feature bank must be stored in memory for inference. To address both issues, we propose an alternative to attention-based schemes which is based on a low-rank approximation of the memory obtained using Singular Value Decomposition. Our scheme has two advantages: (a) it reduces complexity by more than an order of magnitude, and (b) it is amenable to an efficient implementation for the calculation of the memory bases in an incremental fashion which does not require the storage of the whole feature bank in memory. The proposed scheme matches or surpasses the accuracy achieved by attention-based mechanisms while being memory-efficient. Through extensive experiments, we demonstrate that our framework generalises to different architectures and tasks, outperforming the state-of-the-art in three datasets.
Authors: Diwei Sheng, Giles Hamilton-Fletcher, Mahya Beheshti, Chen Feng, John-Ross Rizzo
Abstract: Curbs serve as vital borders that delineate safe pedestrian zones from potential vehicular traffic hazards. Curbs also represent a primary spatial hazard during dynamic navigation with significant stumbling potential. Such vulnerabilities are particularly exacerbated for persons with blindness and low vision (PBLV). Accurate visual-based discrimination of curbs is paramount for assistive technologies that aid PBLV with safe navigation in urban environments. Herein, we investigate the efficacy of curb segmentation for foundation models. We introduce the largest curb segmentation dataset to-date to benchmark leading foundation models. Our results show that state-of-the-art foundation models face significant challenges in curb segmentation. This is due to their high false-positive rates (up to 95%) with poor performance distinguishing curbs from curb-like objects or non-curb areas, such as sidewalks. In addition, the best-performing model averaged a 3.70-second inference time, underscoring problems in providing real-time assistance. In response, we propose solutions including filtered bounding box selections to achieve more accurate curb segmentation. Overall, despite the immediate flexibility of foundation models, their application for practical assistive technology applications still requires refinement. This research highlights the critical need for specialized datasets and tailored model training to address navigation challenges for PBLV and underscores implicit weaknesses in foundation models.
Authors: X. Wang, Siming Fu, Qihan Huang, Wanggui He, Hao Jiang
Abstract: Recent advancements in text-to-image generation models have dramatically enhanced the generation of photorealistic images from textual prompts, leading to an increased interest in personalized text-to-image applications, particularly in multi-subject scenarios. However, these advances are hindered by two main challenges: firstly, the need to accurately maintain the details of each referenced subject in accordance with the textual descriptions; and secondly, the difficulty in achieving a cohesive representation of multiple subjects in a single image without introducing inconsistencies. To address these concerns, our research introduces the MS-Diffusion framework for layout-guided zero-shot image personalization with multi-subjects. This innovative approach integrates grounding tokens with the feature resampler to maintain detail fidelity among subjects. With the layout guidance, MS-Diffusion further improves the cross-attention to adapt to the multi-subject inputs, ensuring that each subject condition acts on specific areas. The proposed multi-subject cross-attention orchestrates harmonious inter-subject compositions while preserving the control of texts. Comprehensive quantitative and qualitative experiments affirm that this method surpasses existing models in both image and text fidelity, promoting the development of personalized text-to-image generation.
Authors: Jie Yang, Bingliang Li, Ailing Zeng, Lei Zhang, Ruimao Zhang
Abstract: In this paper, we develop \textbf{MP-HOI}, a powerful Multi-modal Prompt-based HOI detector designed to leverage both textual descriptions for open-set generalization and visual exemplars for handling high ambiguity in descriptions, realizing HOI detection in the open world. Specifically, it integrates visual prompts into existing language-guided-only HOI detectors to handle situations where textual descriptions face difficulties in generalization and to address complex scenarios with high interaction ambiguity. To facilitate MP-HOI training, we build a large-scale HOI dataset named Magic-HOI, which gathers six existing datasets into a unified label space, forming over 186K images with 2.4K objects, 1.2K actions, and 20K HOI interactions. Furthermore, to tackle the long-tail issue within the Magic-HOI dataset, we introduce an automated pipeline for generating realistically annotated HOI images and present SynHOI, a high-quality synthetic HOI dataset containing 100K images. Leveraging these two datasets, MP-HOI optimizes the HOI task as a similarity learning process between multi-modal prompts and objects/interactions via a unified contrastive loss, to learn generalizable and transferable objects/interactions representations from large-scale data. MP-HOI could serve as a generalist HOI detector, surpassing the HOI vocabulary of existing expert models by more than 30 times. Concurrently, our results demonstrate that MP-HOI exhibits remarkable zero-shot capability in real-world scenarios and consistently achieves a new state-of-the-art performance across various benchmarks.
Authors: Tim Menzner, Jochen L. Leidner, Florian Mittag
Abstract: In this demonstration, we present Country Guesser, a live system that guesses the country that a photo is taken in. In particular, given a Google Street View image, our federated ranking model uses a combination of computer vision, machine learning and text retrieval methods to compute a ranking of likely countries of the location shown in a given image from Street View. Interestingly, using text-based features to probe large pre-trained language models can assist to provide cross-modal supervision. We are not aware of previous country guessing systems informed by visual and textual features.
Authors: Weiyun Wang, Shuibo Zhang, Yiming Ren, Yuchen Duan, Tiantong Li, Shuo Liu, Mengkang Hu, Zhe Chen, Kaipeng Zhang, Lewei Lu, Xizhou Zhu, Ping Luo, Yu Qiao, Jifeng Dai, Wenqi Shao, Wenhai Wang
Abstract: With the rapid advancement of multimodal large language models (MLLMs), their evaluation has become increasingly comprehensive. However, understanding long multimodal content, as a foundational ability for real-world applications, remains underexplored. In this work, we present Needle In A Multimodal Haystack (MM-NIAH), the first benchmark specifically designed to systematically evaluate the capability of existing MLLMs to comprehend long multimodal documents. Our benchmark includes three types of evaluation tasks: multimodal retrieval, counting, and reasoning. In each task, the model is required to answer the questions according to different key information scattered throughout the given multimodal document. Evaluating the leading MLLMs on MM-NIAH, we observe that existing models still have significant room for improvement on these tasks, especially on vision-centric evaluation. We hope this work can provide a platform for further research on long multimodal document comprehension and contribute to the advancement of MLLMs. Code and benchmark are released at https://github.com/OpenGVLab/MM-NIAH.
Authors: Athanasios Tragakis, Marco Aversa, Chaitanya Kaul, Roderick Murray-Smith, Daniele Faccio
Abstract: In this work, we introduce Pixelsmith, a zero-shot text-to-image generative framework to sample images at higher resolutions with a single GPU. We are the first to show that it is possible to scale the output of a pre-trained diffusion model by a factor of 1000, opening the road for gigapixel image generation at no additional cost. Our cascading method uses the image generated at the lowest resolution as a baseline to sample at higher resolutions. For the guidance, we introduce the Slider, a tunable mechanism that fuses the overall structure contained in the first-generated image with enhanced fine details. At each inference step, we denoise patches rather than the entire latent space, minimizing memory demands such that a single GPU can handle the process, regardless of the image's resolution. Our experimental results show that Pixelsmith not only achieves higher quality and diversity compared to existing techniques, but also reduces sampling time and artifacts. The code for our work is available at https://github.com/Thanos-DB/Pixelsmith.
Authors: Long Peng, Wenbo Li, Renjing Pei, Jingjing Ren, Yang Wang, Yang Cao, Zheng-Jun Zha
Abstract: Existing image super-resolution (SR) techniques often fail to generalize effectively in complex real-world settings due to the significant divergence between training data and practical scenarios. To address this challenge, previous efforts have either manually simulated intricate physical-based degradations or utilized learning-based techniques, yet these approaches remain inadequate for producing large-scale, realistic, and diverse data simultaneously. In this paper, we introduce a novel Realistic Decoupled Data Generator (RealDGen), an unsupervised learning data generation framework designed for real-world super-resolution. We meticulously develop content and degradation extraction strategies, which are integrated into a novel content-degradation decoupled diffusion model to create realistic low-resolution images from unpaired real LR and HR images. Extensive experiments demonstrate that RealDGen excels in generating large-scale, high-quality paired data that mirrors real-world degradations, significantly advancing the performance of popular SR models on various real-world benchmarks.
Authors: Marian Longa, Jo\~ao F. Henriques
Abstract: Unsupervised object detection using deep neural networks is typically a difficult problem with few to no guarantees about the learned representation. In this work we present the first unsupervised object detection method that is theoretically guaranteed to recover the true object positions up to quantifiable small shifts. We develop an unsupervised object detection architecture and prove that the learned variables correspond to the true object positions up to small shifts related to the encoder and decoder receptive field sizes, the object sizes, and the widths of the Gaussians used in the rendering process. We perform detailed analysis of how the error depends on each of these variables and perform synthetic experiments validating our theoretical predictions up to a precision of individual pixels. We also perform experiments on CLEVR-based data and show that, unlike current SOTA object detection methods (SAM, CutLER), our method's prediction errors always lie within our theoretical bounds. We hope that this work helps open up an avenue of research into object detection methods with theoretical guarantees.
Authors: Kamil Jeziorek, Piotr Wzorek, Krzysztof Blachut, Andrea Pinna, Tomasz Kryjak
Abstract: The utilisation of event cameras represents an important and swiftly evolving trend aimed at addressing the constraints of traditional video systems. Particularly within the automotive domain, these cameras find significant relevance for their integration into embedded real-time systems due to lower latency and energy consumption. One effective approach to ensure the necessary throughput and latency for event processing systems is through the utilisation of graph convolutional networks (GCNs). In this study, we introduce a series of hardware-aware optimisations tailored for PointNet++, a GCN architecture designed for point cloud processing. The proposed techniques result in more than a 100-fold reduction in model size compared to Asynchronous Event-based GNN (AEGNN), one of the most recent works in the field, with a relatively small decrease in accuracy (2.3% for N-Caltech101 classification, 1.7% for N-Cars classification), thus following the TinyML trend. Based on software research, we designed a custom EFGCN (Event-Based FPGA-accelerated Graph Convolutional Network) and we implemented it on ZCU104 SoC FPGA platform, achieving a throughput of 13.3 million events per second (MEPS) and real-time partially asynchronous processing with a latency of 4.47 ms. We also address the scalability of the proposed hardware model to improve the obtained accuracy score. To the best of our knowledge, this study marks the first endeavour in accelerating PointNet++ networks on SoC FPGAs, as well as the first hardware architecture exploration of graph convolutional networks implementation for real-time continuous event data processing. We publish both software and hardware source code in an open repository: https://github.com/vision-agh/*** (will be published upon acceptance).
Authors: Riccardo Fogliato, Pratik Patil, Mathew Monfort, Pietro Perona
Abstract: Model performance evaluation is a critical and expensive task in machine learning and computer vision. Without clear guidelines, practitioners often estimate model accuracy using a one-time random selection of the data. However, by employing tailored sampling and estimation strategies, one can obtain more precise estimates and reduce annotation costs. In this paper, we propose a statistical framework for model evaluation that includes stratification, sampling, and estimation components. We examine the statistical properties of each component and evaluate their efficiency (precision). One key result of our work is that stratification via k-means clustering based on accurate predictions of model performance yields efficient estimators. Our experiments on computer vision datasets show that this method consistently provides more precise accuracy estimates than the traditional simple random sampling, even with substantial efficiency gains of 10x. We also find that model-assisted estimators, which leverage predictions of model accuracy on the unlabeled portion of the dataset, are generally more efficient than the traditional estimates based solely on the labeled data.
Authors: Chao Wang, Krzysztof Wolski, Bernhard Kerbl, Ana Serrano, Mojtaba Bemana, Hans-Peter Seidel, Karol Myszkowski, Thomas Leimk\"uhler
Abstract: Radiance field methods represent the state of the art in reconstructing complex scenes from multi-view photos. However, these reconstructions often suffer from one or both of the following limitations: First, they typically represent scenes in low dynamic range (LDR), which restricts their use to evenly lit environments and hinders immersive viewing experiences. Secondly, their reliance on a pinhole camera model, assuming all scene elements are in focus in the input images, presents practical challenges and complicates refocusing during novel-view synthesis. Addressing these limitations, we present a lightweight method based on 3D Gaussian Splatting that utilizes multi-view LDR images of a scene with varying exposure times, apertures, and focus distances as input to reconstruct a high-dynamic-range (HDR) radiance field. By incorporating analytical convolutions of Gaussians based on a thin-lens camera model as well as a tonemapping module, our reconstructions enable the rendering of HDR content with flexible refocusing capabilities. We demonstrate that our combined treatment of HDR and depth of field facilitates real-time cinematic rendering, outperforming the state of the art.
Authors: Challapalli Phanindra Revanth, Sumohana S. Channappayya, C Krishna Mohan
Abstract: Computing the loss gradient via backpropagation consumes considerable energy during deep learning (DL) model training. In this paper, we propose a novel approach to efficiently compute DL models' gradients to mitigate the substantial energy overhead associated with backpropagation. Exploiting the over-parameterized nature of DL models and the smoothness of their loss landscapes, we propose a method called {\em GradSamp} for sampling gradient updates from a Gaussian distribution. Specifically, we update model parameters at a given epoch (chosen periodically or randomly) by perturbing the parameters (element-wise) from the previous epoch with Gaussian ``noise''. The parameters of the Gaussian distribution are estimated using the error between the model parameter values from the two previous epochs. {\em GradSamp} not only streamlines gradient computation but also enables skipping entire epochs, thereby enhancing overall efficiency. We rigorously validate our hypothesis across a diverse set of standard and non-standard CNN and transformer-based models, spanning various computer vision tasks such as image classification, object detection, and image segmentation. Additionally, we explore its efficacy in out-of-distribution scenarios such as Domain Adaptation (DA), Domain Generalization (DG), and decentralized settings like Federated Learning (FL). Our experimental results affirm the effectiveness of {\em GradSamp} in achieving notable energy savings without compromising performance, underscoring its versatility and potential impact in practical DL applications.
Authors: Haiming Yao, Wei Luo, Yunkang Cao, Yiheng Zhang, Wenyong Yu, Weiming Shen
Abstract: Texture surface anomaly detection finds widespread applications in industrial settings. However, existing methods often necessitate gathering numerous samples for model training. Moreover, they predominantly operate within a close-set detection framework, limiting their ability to identify anomalies beyond the training dataset. To tackle these challenges, this paper introduces a novel zero-shot texture anomaly detection method named Global-Regularized Neighborhood Regression (GRNR). Unlike conventional approaches, GRNR can detect anomalies on arbitrary textured surfaces without any training data or cost. Drawing from human visual cognition, GRNR derives two intrinsic prior supports directly from the test texture image: local neighborhood priors characterized by coherent similarities and global normality priors featuring typical normal patterns. The fundamental principle of GRNR involves utilizing the two extracted intrinsic support priors for self-reconstructive regression of the query sample. This process employs the transformation facilitated by local neighbor support while being regularized by global normality support, aiming to not only achieve visually consistent reconstruction results but also preserve normality properties. We validate the effectiveness of GRNR across various industrial scenarios using eight benchmark datasets, demonstrating its superior detection performance without the need for training data. Remarkably, our method is applicable for open-set texture defect detection and can even surpass existing vanilla approaches that require extensive training.
Authors: Rohit Jena, Pratik Chaudhari, James C. Gee
Abstract: Deep Learning in Image Registration (DLIR) methods have been tremendously successful in image registration due to their speed and ability to incorporate weak label supervision at training time. However, DLIR methods forego many of the benefits of classical optimization-based methods. The functional nature of deep networks do not guarantee that the predicted transformation is a local minima of the registration objective, the representation of the transformation (displacement/velocity field/affine) is fixed, and the networks are not robust to domain shift. Our method aims to bridge this gap between classical and learning methods by incorporating optimization as a layer in a deep network. A deep network is trained to predict multi-scale dense feature images that are registered using a black box iterative optimization solver. This optimal warp is then used to minimize image and label alignment errors. By implicitly differentiating end-to-end through an iterative optimization solver, our learned features are registration and label-aware, and the warp functions are guaranteed to be local minima of the registration objective in the feature space. Our framework shows excellent performance on in-domain datasets, and is agnostic to domain shift such as anisotropy and varying intensity profiles. For the first time, our method allows switching between arbitrary transformation representations (free-form to diffeomorphic) at test time with zero retraining. End-to-end feature learning also facilitates interpretability of features, and out-of-the-box promptability using additional label-fidelity terms at inference.
Authors: Huiwon Jang, Dongyoung Kim, Junsu Kim, Jinwoo Shin, Pieter Abbeel, Younggyo Seo
Abstract: Self-supervised learning of image representations by predicting future frames is a promising direction but still remains a challenge. This is because of the under-determined nature of frame prediction; multiple potential futures can arise from a single current frame. To tackle this challenge, in this paper, we revisit the idea of stochastic video generation that learns to capture uncertainty in frame prediction and explore its effectiveness for representation learning. Specifically, we design a framework that trains a stochastic frame prediction model to learn temporal information between frames. Moreover, to learn dense information within each frame, we introduce an auxiliary masked image modeling objective along with a shared decoder architecture. We find this architecture allows for combining both objectives in a synergistic and compute-efficient manner. We demonstrate the effectiveness of our framework on a variety of tasks from video label propagation and vision-based robot learning domains, such as video segmentation, pose tracking, vision-based robotic locomotion, and manipulation tasks. Code is available on the project webpage: https://sites.google.com/view/2024rsp.
Authors: Shashank Agnihotri, Julia Grabinski, Janis Keuper, Margret Keuper
Abstract: Image restoration networks are usually comprised of an encoder and a decoder, responsible for aggregating image content from noisy, distorted data and to restore clean, undistorted images, respectively. Data aggregation as well as high-resolution image generation both usually come at the risk of involving aliases, i.e.~standard architectures put their ability to reconstruct the model input in jeopardy to reach high PSNR values on validation data. The price to be paid is low model robustness. In this work, we show that simply providing alias-free paths in state-of-the-art reconstruction transformers supports improved model robustness at low costs on the restoration performance. We do so by proposing BOA-Restormer, a transformer-based image restoration model that executes downsampling and upsampling operations partly in the frequency domain to ensure alias-free paths along the entire model while potentially preserving all relevant high-frequency information.
Authors: Shuvendu Roy, Yasaman Parhizkar, Franklin Ogidi, Vahid Reza Khazaie, Michael Colacci, Ali Etemad, Elham Dolatabadi, Arash Afkanpour
Abstract: We perform a comprehensive benchmarking of contrastive frameworks for learning multimodal representations in the medical domain. Through this study, we aim to answer the following research questions: (i) How transferable are general-domain representations to the medical domain? (ii) Is multimodal contrastive training sufficient, or does it benefit from unimodal training as well? (iii) What is the impact of feature granularity on the effectiveness of multimodal medical representation learning? To answer these questions, we investigate eight contrastive learning approaches under identical training setups, and train them on 2.8 million image-text pairs from four datasets, and evaluate them on 25 downstream tasks, including classification (zero-shot and linear probing), image-to-text and text-to-image retrieval, and visual question-answering. Our findings suggest a positive answer to the first question, a negative answer to the second question, and the benefit of learning fine-grained features. Finally, we make our code publicly available.
Authors: Ming Hu, Peng Xia, Lin Wang, Siyuan Yan, Feilong Tang, Zhongxing Xu, Yimin Luo, Kaimin Song, Jurgen Leitner, Xuelian Cheng, Jun Cheng, Chi Liu, Kaijing Zhou, Zongyuan Ge
Abstract: Surgical scene perception via videos are critical for advancing robotic surgery, telesurgery, and AI-assisted surgery, particularly in ophthalmology. However, the scarcity of diverse and richly annotated video datasets has hindered the development of intelligent systems for surgical workflow analysis. Existing datasets for surgical workflow analysis, which typically face challenges such as small scale, a lack of diversity in surgery and phase categories, and the absence of time-localized annotations, limit the requirements for action understanding and model generalization validation in complex and diverse real-world surgical scenarios. To address this gap, we introduce OphNet, a large-scale, expert-annotated video benchmark for ophthalmic surgical workflow understanding. OphNet features: 1) A diverse collection of 2,278 surgical videos spanning 66 types of cataract, glaucoma, and corneal surgeries, with detailed annotations for 102 unique surgical phases and 150 granular operations; 2) It offers sequential and hierarchical annotations for each surgery, phase, and operation, enabling comprehensive understanding and improved interpretability; 3) Moreover, OphNet provides time-localized annotations, facilitating temporal localization and prediction tasks within surgical workflows. With approximately 205 hours of surgical videos, OphNet is about 20 times larger than the largest existing surgical workflow analysis benchmark. Our dataset and code have been made available at: \url{https://github.com/minghu0830/OphNet-benchmark}.
Authors: Heng Yu, Chaoyang Wang, Peiye Zhuang, Willi Menapace, Aliaksandr Siarohin, Junli Cao, Laszlo A Jeni, Sergey Tulyakov, Hsin-Ying Lee
Abstract: Existing dynamic scene generation methods mostly rely on distilling knowledge from pre-trained 3D generative models, which are typically fine-tuned on synthetic object datasets. As a result, the generated scenes are often object-centric and lack photorealism. To address these limitations, we introduce a novel pipeline designed for photorealistic text-to-4D scene generation, discarding the dependency on multi-view generative models and instead fully utilizing video generative models trained on diverse real-world datasets. Our method begins by generating a reference video using the video generation model. We then learn the canonical 3D representation of the video using a freeze-time video, delicately generated from the reference video. To handle inconsistencies in the freeze-time video, we jointly learn a per-frame deformation to model these imperfections. We then learn the temporal deformation based on the canonical representation to capture dynamic interactions in the reference video. The pipeline facilitates the generation of dynamic scenes with enhanced photorealism and structural integrity, viewable from multiple perspectives, thereby setting a new standard in 4D scene generation.
Authors: Zesen Cheng, Sicong Leng, Hang Zhang, Yifei Xin, Xin Li, Guanzheng Chen, Yongxin Zhu, Wenqi Zhang, Ziyang Luo, Deli Zhao, Lidong Bing
Abstract: In this paper, we present the VideoLLaMA 2, a set of Video Large Language Models (Video-LLMs) designed to enhance spatial-temporal modeling and audio understanding in video and audio-oriented tasks. Building upon its predecessor, VideoLLaMA 2 incorporates a tailor-made Spatial-Temporal Convolution (STC) connector, which effectively captures the intricate spatial and temporal dynamics of video data. Additionally, we integrate an Audio Branch into the model through joint training, thereby enriching the multimodal understanding capabilities of the model by seamlessly incorporating audio cues. Comprehensive evaluations on multiple-choice video question answering (MC-VQA), open-ended video question answering (OE-VQA), and video captioning (VC) tasks demonstrate that VideoLLaMA 2 consistently achieves competitive results among open-source models and even gets close to some proprietary models on several benchmarks. Furthermore, VideoLLaMA 2 exhibits reasonable improvements in audio-only and audio-video question-answering (AQA & OE-AVQA) benchmarks over existing models. These advancements underline VideoLLaMA 2's superior performance in multimodal comprehension, setting a new standard for intelligent video analysis systems. All models are public to facilitate further research.
Authors: Yinbo Chen, Oliver Wang, Richard Zhang, Eli Shechtman, Xiaolong Wang, Michael Gharbi
Abstract: Diffusion models have shown an impressive ability to model complex data distributions, with several key advantages over GANs, such as stable training, better coverage of the training distribution's modes, and the ability to solve inverse problems without extra training. However, most diffusion models learn the distribution of fixed-resolution images. We propose to learn the distribution of continuous images by training diffusion models on image neural fields, which can be rendered at any resolution, and show its advantages over fixed-resolution models. To achieve this, a key challenge is to obtain a latent space that represents photorealistic image neural fields. We propose a simple and effective method, inspired by several recent techniques but with key changes to make the image neural fields photorealistic. Our method can be used to convert existing latent diffusion autoencoders into image neural field autoencoders. We show that image neural field diffusion models can be trained using mixed-resolution image datasets, outperform fixed-resolution diffusion models followed by super-resolution models, and can solve inverse problems with conditions applied at different scales efficiently.
Authors: Biplov Bhandari, Timothy Mayer
Abstract: The Bhutanese government is increasing its utilization of technological approaches such as including Remote Sensing-based knowledge in their decision-making process. This study focuses on crop type and crop extent in Paro, one of the top rice-yielding districts in Bhutan, and employs publicly available NICFI high-resolution satellite imagery from Planet. Two Deep Learning (DL) approaches, point-based (DNN) and patch-based (U-Net), models were used in conjunction with cloud-computing platforms. Three different models per DL approaches (DNN and U-Net) were trained: 1) RGBN channels from Planet; 2) RGBN and elevation data (RGBNE); 3) RGBN and Sentinel-1 (S1) data (RGBNS), and RGBN with E and S1 data (RGBNES). From this comprehensive analysis, the U-Net displayed higher performance metrics across both model training and model validation efforts. Among the U-Net model sets, the RGBN, RGBNE, RGBNS, and RGBNES models had an F1-score of 0.8546, 0.8563, 0.8467, and 0.8500 respectively. An independent model evaluation was performed and found a high level of performance variation across all the metrics. For this independent model evaluation, the U-Net RGBN, RGBNE, RGBNES, and RGBN models displayed the F1-scores of 0.5935, 0.6154, 0.5882, and 0.6582, suggesting U-Net RGBNES as the best model. The study shows that the DL approaches can predict rice. Also, DL methods can be used with the survey-based approaches currently utilized by the Bhutan Department of Agriculture. Further, this study demonstrated the usage of regional land cover products such as SERVIR's RLCMS as a weak label approach to capture different strata addressing the class imbalance problem and improving the sampling design for DL application. Finally, through preliminary model testing and comparisons outlined it was shown that using additional features such as NDVI, EVI, and NDWI did not drastically improve model performance.
Authors: Hang Yao, Ming Liu, Haolin Wang, Zhicun Yin, Zifei Yan, Xiaopeng Hong, Wangmeng Zuo
Abstract: Diffusion models have shown superior performance on unsupervised anomaly detection tasks. Since trained with normal data only, diffusion models tend to reconstruct normal counterparts of test images with certain noises added. However, these methods treat all potential anomalies equally, which may cause two main problems. From the global perspective, the difficulty of reconstructing images with different anomalies is uneven. Therefore, instead of utilizing the same setting for all samples, we propose to predict a particular denoising step for each sample by evaluating the difference between image contents and the priors extracted from diffusion models. From the local perspective, reconstructing abnormal regions differs from normal areas even in the same image. Theoretically, the diffusion model predicts a noise for each step, typically following a standard Gaussian distribution. However, due to the difference between the anomaly and its potential normal counterpart, the predicted noise in abnormal regions will inevitably deviate from the standard Gaussian distribution. To this end, we propose introducing synthetic abnormal samples in training to encourage the diffusion models to break through the limitation of standard Gaussian distribution, and a spatial-adaptive feature fusion scheme is utilized during inference. With the above modifications, we propose a global and local adaptive diffusion model (abbreviated to GLAD) for unsupervised anomaly detection, which introduces appealing flexibility and achieves anomaly-free reconstruction while retaining as much normal information as possible. Extensive experiments are conducted on three commonly used anomaly detection datasets (MVTec-AD, MPDD, and VisA) and a printed circuit board dataset (PCB-Bank) we integrated, showing the effectiveness of the proposed method.
Authors: John Yang, Le An, Su Inn Park
Abstract: Transformers have excelled in many tasks including vision. However, efficient deployment of transformer models in low-latency or high-throughput applications is hindered by the computation in the attention mechanism which involves expensive operations such as matrix multiplication and Softmax. To address this, we introduce ReduceFormer, a family of models optimized for efficiency with the spirit of attention. ReduceFormer leverages only simple operations such as reduction and element-wise multiplication, leading to greatly simplified architecture and improved inference performance, with up to 37% reduction in latency and 44% improvement in throughput, while maintaining competitive accuracy comparable to other recent methods. The proposed model family is suitable for edge devices where compute resource and memory bandwidth are limited, as well as for cloud computing where high throughput is sought after.
Authors: Lue Fan, Yuxue Yang, Minxing Li, Hongsheng Li, Zhaoxiang Zhang
Abstract: In this paper, we introduce Trim 3D Gaussian Splatting (TrimGS) to reconstruct accurate 3D geometry from images. Previous arts for geometry reconstruction from 3D Gaussians mainly focus on exploring strong geometry regularization. Instead, from a fresh perspective, we propose to obtain accurate 3D geometry of a scene by Gaussian trimming, which selectively removes the inaccurate geometry while preserving accurate structures. To achieve this, we analyze the contributions of individual 3D Gaussians and propose a contribution-based trimming strategy to remove the redundant or inaccurate Gaussians. Furthermore, our experimental and theoretical analyses reveal that a relatively small Gaussian scale is a non-negligible factor in representing and optimizing the intricate details. Therefore the proposed TrimGS maintains relatively small Gaussian scales. In addition, TrimGS is also compatible with the effective geometry regularization strategies in previous arts. When combined with the original 3DGS and the state-of-the-art 2DGS, TrimGS consistently yields more accurate geometry and higher perceptual quality. Our project page is https://trimgs.github.io
URLs: https://trimgs.github.io
Authors: Javier Montalvo, Juan Ignacio Bravo P\'erez-Villar, \'Alvaro Garc\'ia-Mart\'in, Pablo Carballeira, Jes\'us Besc'os
Abstract: Data acquired in space operational conditions is scarce due to the costs and complexity of space operations. This poses a challenge to learning-based visual-based navigation algorithms employed in autonomous spacecraft navigation. Existing datasets, which largely depend on computer-simulated data, have partially filled this gap. However, the image generation tools they use are proprietary, which limits the evaluation of methods to unseen scenarios. Furthermore, these datasets provide limited ground-truth data, primarily focusing on the spacecraft's translation and rotation relative to the camera. To address these limitations, we present SPIN (SPacecraft Imagery for Navigation), an open-source realistic spacecraft image generation tool for relative navigation between two spacecrafts. SPIN provides a wide variety of ground-truth data and allows researchers to employ custom 3D models of satellites, define specific camera-relative poses, and adjust various settings such as camera parameters and environmental illumination conditions. For the task of spacecraft pose estimation, we compare the results of training with a SPIN-generated dataset against existing synthetic datasets. We show a %50 average error reduction in common testbed data (that simulates realistic space conditions). Both the SPIN tool (and source code) and our enhanced version of the synthetic datasets will be publicly released upon paper acceptance on GitHub https://github.com/vpulab/SPIN.
Authors: Renjie Pi, Jianshu Zhang, Jipeng Zhang, Rui Pan, Zhekai Chen, Tong Zhang
Abstract: Image description datasets play a crucial role in the advancement of various applications such as image understanding, text-to-image generation, and text-image retrieval. Currently, image description datasets primarily originate from two sources. One source is the scraping of image-text pairs from the web. Despite their abundance, these descriptions are often of low quality and noisy. Another is through human labeling. Datasets such as COCO are generally very short and lack details. Although detailed image descriptions can be annotated by humans, the high annotation cost limits the feasibility. These limitations underscore the need for more efficient and scalable methods to generate accurate and detailed image descriptions. In this paper, we propose an innovative framework termed Image Textualization (IT), which automatically produces high-quality image descriptions by leveraging existing multi-modal large language models (MLLMs) and multiple vision expert models in a collaborative manner, which maximally convert the visual information into text. To address the current lack of benchmarks for detailed descriptions, we propose several benchmarks for comprehensive evaluation, which verifies the quality of image descriptions created by our framework. Furthermore, we show that LLaVA-7B, benefiting from training on IT-curated descriptions, acquire improved capability to generate richer image descriptions, substantially increasing the length and detail of their output with less hallucination.
Authors: Brandon Trabucco, Max Gurinas, Kyle Doherty, Ruslan Salakhutdinov
Abstract: Large multimodal models such as Stable Diffusion can generate, detect, and classify new visual concepts after fine-tuning just a single word embedding. Do models learn similar words for the same concepts (i.e.
Authors: Nikos Kolotouros, Thiemo Alldieck, Enric Corona, Eduard Gabriel Bazavan, Cristian Sminchisescu
Abstract: We present AvatarPopUp, a method for fast, high quality 3D human avatar generation from different input modalities, such as images and text prompts and with control over the generated pose and shape. The common theme is the use of diffusion-based image generation networks that are specialized for each particular task, followed by a 3D lifting network. We purposefully decouple the generation from the 3D modeling which allow us to leverage powerful image synthesis priors, trained on billions of text-image pairs. We fine-tune latent diffusion networks with additional image conditioning to solve tasks such as image generation and back-view prediction, and to support qualitatively different multiple 3D hypotheses. Our partial fine-tuning approach allows to adapt the networks for each task without inducing catastrophic forgetting. In our experiments, we demonstrate that our method produces accurate, high-quality 3D avatars with diverse appearance that respect the multimodal text, image, and body control signals. Our approach can produce a 3D model in as few as 2 seconds, a four orders of magnitude speedup w.r.t. the vast majority of existing methods, most of which solve only a subset of our tasks, and with fewer controls, thus enabling applications that require the controlled 3D generation of human avatars at scale. The project website can be found at https://www.nikoskolot.com/avatarpopup/.
Authors: Haian Jin, Yuan Li, Fujun Luan, Yuanbo Xiangli, Sai Bi, Kai Zhang, Zexiang Xu, Jin Sun, Noah Snavely
Abstract: Single-image relighting is a challenging task that involves reasoning about the complex interplay between geometry, materials, and lighting. Many prior methods either support only specific categories of images, such as portraits, or require special capture conditions, like using a flashlight. Alternatively, some methods explicitly decompose a scene into intrinsic components, such as normals and BRDFs, which can be inaccurate or under-expressive. In this work, we propose a novel end-to-end 2D relighting diffusion model, called Neural Gaffer, that takes a single image of any object and can synthesize an accurate, high-quality relit image under any novel environmental lighting condition, simply by conditioning an image generator on a target environment map, without an explicit scene decomposition. Our method builds on a pre-trained diffusion model, and fine-tunes it on a synthetic relighting dataset, revealing and harnessing the inherent understanding of lighting present in the diffusion model. We evaluate our model on both synthetic and in-the-wild Internet imagery and demonstrate its advantages in terms of generalization and accuracy. Moreover, by combining with other generative methods, our model enables many downstream 2D tasks, such as text-based relighting and object insertion. Our model can also operate as a strong relighting prior for 3D tasks, such as relighting a radiance field.
Authors: Sucheng Ren, Xianhang Li, Haoqin Tu, Feng Wang, Fangxun Shu, Lei Zhang, Jieru Mei, Linjie Yang, Peng Wang, Heng Wang, Alan Yuille, Cihang Xie
Abstract: The vision community has started to build with the recently developed state space model, Mamba, as the new backbone for a range of tasks. This paper shows that Mamba's visual capability can be significantly enhanced through autoregressive pretraining, a direction not previously explored. Efficiency-wise, the autoregressive nature can well capitalize on the Mamba's unidirectional recurrent structure, enabling faster overall training speed compared to other training strategies like mask modeling. Performance-wise, autoregressive pretraining equips the Mamba architecture with markedly higher accuracy over its supervised-trained counterparts and, more importantly, successfully unlocks its scaling potential to large and even huge model sizes. For example, with autoregressive pretraining, a base-size Mamba attains 83.2\% ImageNet accuracy, outperforming its supervised counterpart by 2.0\%; our huge-size Mamba, the largest Vision Mamba to date, attains 85.0\% ImageNet accuracy (85.5\% when finetuned with $384\times384$ inputs), notably surpassing all other Mamba variants in vision. The code is available at \url{https://github.com/OliverRensu/ARM}.
Authors: Kuan Heng Lin, Sicheng Mo, Ben Klingher, Fangzhou Mu, Bolei Zhou
Abstract: Recent controllable generation approaches such as FreeControl and Diffusion Self-guidance bring fine-grained spatial and appearance control to text-to-image (T2I) diffusion models without training auxiliary modules. However, these methods optimize the latent embedding for each type of score function with longer diffusion steps, making the generation process time-consuming and limiting their flexibility and use. This work presents Ctrl-X, a simple framework for T2I diffusion controlling structure and appearance without additional training or guidance. Ctrl-X designs feed-forward structure control to enable the structure alignment with a structure image and semantic-aware appearance transfer to facilitate the appearance transfer from a user-input image. Extensive qualitative and quantitative experiments illustrate the superior performance of Ctrl-X on various condition inputs and model checkpoints. In particular, Ctrl-X supports novel structure and appearance control with arbitrary condition images of any modality, exhibits superior image quality and appearance transfer compared to existing works, and provides instant plug-and-play functionality to any T2I and text-to-video (T2V) diffusion model. See our project page for an overview of the results: https://genforce.github.io/ctrl-x
Authors: Chenyu Yang, Xizhou Zhu, Jinguo Zhu, Weijie Su, Junjie Wang, Xuan Dong, Wenhai Wang, Lewei Lu, Bin Li, Jie Zhou, Yu Qiao, Jifeng Dai
Abstract: Recently, vision model pre-training has evolved from relying on manually annotated datasets to leveraging large-scale, web-crawled image-text data. Despite these advances, there is no pre-training method that effectively exploits the interleaved image-text data, which is very prevalent on the Internet. Inspired by the recent success of compression learning in natural language processing, we propose a novel vision model pre-training method called Latent Compression Learning (LCL) for interleaved image-text data. This method performs latent compression learning by maximizing the mutual information between the inputs and outputs of a causal attention model. The training objective can be decomposed into two basic tasks: 1) contrastive learning between visual representation and preceding context, and 2) generating subsequent text based on visual representation. Our experiments demonstrate that our method not only matches the performance of CLIP on paired pre-training datasets (e.g., LAION), but can also leverage interleaved pre-training data (e.g., MMC4) to learn robust visual representation from scratch, showcasing the potential of vision model pre-training with interleaved image-text data. Code is released at https://github.com/OpenGVLab/LCL.
Authors: Yunze Man, Liang-Yan Gui, Yu-Xiong Wang
Abstract: Being able to carry out complicated vision language reasoning tasks in 3D space represents a significant milestone in developing household robots and human-centered embodied AI. In this work, we demonstrate that a critical and distinct challenge in 3D vision language reasoning is situational awareness, which incorporates two key components: (1) The autonomous agent grounds its self-location based on a language prompt. (2) The agent answers open-ended questions from the perspective of its calculated position. To address this challenge, we introduce SIG3D, an end-to-end Situation-Grounded model for 3D vision language reasoning. We tokenize the 3D scene into sparse voxel representation and propose a language-grounded situation estimator, followed by a situated question answering module. Experiments on the SQA3D and ScanQA datasets show that SIG3D outperforms state-of-the-art models in situation estimation and question answering by a large margin (e.g., an enhancement of over 30% on situation estimation accuracy). Subsequent analysis corroborates our architectural design choices, explores the distinct functions of visual and textual tokens, and highlights the importance of situational awareness in the domain of 3D question answering.
Authors: Xingyu Fu, Muyu He, Yujie Lu, William Yang Wang, Dan Roth
Abstract: We present a novel task and benchmark for evaluating the ability of text-to-image(T2I) generation models to produce images that fit commonsense in real life, which we call Commonsense-T2I. Given two adversarial text prompts containing an identical set of action words with minor differences, such as "a lightbulb without electricity" v.s. "a lightbulb with electricity", we evaluate whether T2I models can conduct visual-commonsense reasoning, e.g. produce images that fit "the lightbulb is unlit" vs. "the lightbulb is lit" correspondingly. Commonsense-T2I presents an adversarial challenge, providing pairwise text prompts along with expected outputs. The dataset is carefully hand-curated by experts and annotated with fine-grained labels, such as commonsense type and likelihood of the expected outputs, to assist analyzing model behavior. We benchmark a variety of state-of-the-art (sota) T2I models and surprisingly find that, there is still a large gap between image synthesis and real life photos--even the DALL-E 3 model could only achieve 48.92% on Commonsense-T2I, and the stable diffusion XL model only achieves 24.92% accuracy. Our experiments show that GPT-enriched prompts cannot solve this challenge, and we include a detailed analysis about possible reasons for such deficiency. We aim for Commonsense-T2I to serve as a high-quality evaluation benchmark for T2I commonsense checking, fostering advancements in real life image generation.
Authors: Xi Chen, Yutong Feng, Mengting Chen, Yiyang Wang, Shilong Zhang, Yu Liu, Yujun Shen, Hengshuang Zhao
Abstract: Image editing serves as a practical yet challenging task considering the diverse demands from users, where one of the hardest parts is to precisely describe how the edited image should look like. In this work, we present a new form of editing, termed imitative editing, to help users exercise their creativity more conveniently. Concretely, to edit an image region of interest, users are free to directly draw inspiration from some in-the-wild references (e.g., some relative pictures come across online), without having to cope with the fit between the reference and the source. Such a design requires the system to automatically figure out what to expect from the reference to perform the editing. For this purpose, we propose a generative training framework, dubbed MimicBrush, which randomly selects two frames from a video clip, masks some regions of one frame, and learns to recover the masked regions using the information from the other frame. That way, our model, developed from a diffusion prior, is able to capture the semantic correspondence between separate images in a self-supervised manner. We experimentally show the effectiveness of our method under various test cases as well as its superiority over existing alternatives. We also construct a benchmark to facilitate further research.
Authors: Yue Zhao, Yuanjun Xiong, Philipp Kr\"ahenb\"uhl
Abstract: We propose a new transformer-based image and video tokenizer with Binary Spherical Quantization (BSQ). BSQ projects the high-dimensional visual embedding to a lower-dimensional hypersphere and then applies binary quantization. BSQ is (1) parameter-efficient without an explicit codebook, (2) scalable to arbitrary token dimensions, and (3) compact: compressing visual data by up to 100$\times$ with minimal distortion. Our tokenizer uses a transformer encoder and decoder with simple block-wise causal masking to support variable-length videos as input. The resulting BSQ-ViT achieves state-of-the-art visual reconstruction quality on image and video reconstruction benchmarks with 2.4$\times$ throughput compared to the best prior methods. Furthermore, by learning an autoregressive prior for adaptive arithmetic coding, BSQ-ViT achieves comparable results on video compression with state-of-the-art video compression standards. BSQ-ViT also enables masked language models to achieve competitive image synthesis quality to GAN- and diffusion-based methods.
Authors: Qihang Yu, Mark Weber, Xueqing Deng, Xiaohui Shen, Daniel Cremers, Liang-Chieh Chen
Abstract: Recent advancements in generative models have highlighted the crucial role of image tokenization in the efficient synthesis of high-resolution images. Tokenization, which transforms images into latent representations, reduces computational demands compared to directly processing pixels and enhances the effectiveness and efficiency of the generation process. Prior methods, such as VQGAN, typically utilize 2D latent grids with fixed downsampling factors. However, these 2D tokenizations face challenges in managing the inherent redundancies present in images, where adjacent regions frequently display similarities. To overcome this issue, we introduce Transformer-based 1-Dimensional Tokenizer (TiTok), an innovative approach that tokenizes images into 1D latent sequences. TiTok provides a more compact latent representation, yielding substantially more efficient and effective representations than conventional techniques. For example, a 256 x 256 x 3 image can be reduced to just 32 discrete tokens, a significant reduction from the 256 or 1024 tokens obtained by prior methods. Despite its compact nature, TiTok achieves competitive performance to state-of-the-art approaches. Specifically, using the same generator framework, TiTok attains 1.97 gFID, outperforming MaskGIT baseline significantly by 4.21 at ImageNet 256 x 256 benchmark. The advantages of TiTok become even more significant when it comes to higher resolution. At ImageNet 512 x 512 benchmark, TiTok not only outperforms state-of-the-art diffusion model DiT-XL/2 (gFID 2.74 vs. 3.04), but also reduces the image tokens by 64x, leading to 410x faster generation process. Our best-performing variant can significantly surpasses DiT-XL/2 (gFID 2.13 vs. 3.04) while still generating high-quality samples 74x faster.
Authors: Huicong Zhang, Haozhe Xie, Hongxun Yao
Abstract: Video deblurring relies on leveraging information from other frames in the video sequence to restore the blurred regions in the current frame. Mainstream approaches employ bidirectional feature propagation, spatio-temporal transformers, or a combination of both to extract information from the video sequence. However, limitations in memory and computational resources constraints the temporal window length of the spatio-temporal transformer, preventing the extraction of longer temporal contextual information from the video sequence. Additionally, bidirectional feature propagation is highly sensitive to inaccurate optical flow in blurry frames, leading to error accumulation during the propagation process. To address these issues, we propose \textbf{BSSTNet}, \textbf{B}lur-aware \textbf{S}patio-temporal \textbf{S}parse \textbf{T}ransformer Network. It introduces the blur map, which converts the originally dense attention into a sparse form, enabling a more extensive utilization of information throughout the entire video sequence. Specifically, BSSTNet (1) uses a longer temporal window in the transformer, leveraging information from more distant frames to restore the blurry pixels in the current frame. (2) introduces bidirectional feature propagation guided by blur maps, which reduces error accumulation caused by the blur frame. The experimental results demonstrate the proposed BSSTNet outperforms the state-of-the-art methods on the GoPro and DVD datasets.
Authors: Ivan Iliash (Technical University of Munich), Simeon Allmendinger (University of Bayreuth), Felix Meissen (Technical University of Munich), Niklas K\"uhl (University of Bayreuth), Daniel R\"uckert (Technical University of Munich)
Abstract: Generative AI, in general, and synthetic visual data generation, in specific, hold much promise for benefiting surgical training by providing photorealism to simulation environments. Current training methods primarily rely on reading materials and observing live surgeries, which can be time-consuming and impractical. In this work, we take a significant step towards improving the training process. Specifically, we use diffusion models in combination with a zero-shot video diffusion method to interactively generate realistic laparoscopic images and videos by specifying a surgical action through text and guiding the generation with tool positions through segmentation masks. We demonstrate the performance of our approach using the publicly available Cholec dataset family and evaluate the fidelity and factual correctness of our generated images using a surgical action recognition model as well as the pixel-wise F1-score for the spatial control of tool generation. We achieve an FID of 38.097 and an F1-score of 0.71.
Authors: Xiaofeng Zhang, Chen Shen, Xiaosong Yuan, Shaotian Yan, Liang Xie, Wenxiao Wang, Chaochen Gu, Hao Tang, Jieping Ye
Abstract: Recently, multimodal large language models have exploded with an endless variety, most of the popular Large Vision Language Models (LVLMs) depend on sequential visual representation, where images are converted into hundreds or thousands of tokens before being input into the Large Language Model (LLM) along with language prompts. The black-box design hinders the interpretability of visual-language models, especially regarding more complex reasoning tasks. To explore the interaction process between image and text in complex reasoning tasks, we introduce the information flow method to visualize the interaction mechanism. By analyzing the dynamic flow of the information flow, we find that the information flow appears to converge in the shallow layer. Further investigation revealed a redundancy of the image token in the shallow layer. Consequently, a truncation strategy was introduced to aggregate image tokens within these shallow layers. This approach has been validated through experiments across multiple models, yielding consistent improvements.
Authors: Lingfeng Zhang, Yuheng Guo, Yepeng Ding, Hiroyuki Sato
Abstract: Online signature verification plays a pivotal role in security infrastructures. However, conventional online signature verification models pose significant risks to data privacy, especially during training processes. To mitigate these concerns, we propose a novel federated learning framework that leverages 1-D Convolutional Neural Networks (CNN) for online signature verification. Furthermore, our experiments demonstrate the effectiveness of our framework regarding 1-D CNN and federated learning. Particularly, the experiment results highlight that our framework 1) minimizes local computational resources; 2) enhances transfer effects with substantial initialization data; 3) presents remarkable scalability. The centralized 1-D CNN model achieves an Equal Error Rate (EER) of 3.33% and an accuracy of 96.25%. Meanwhile, configurations with 2, 5, and 10 agents yield EERs of 5.42%, 5.83%, and 5.63%, along with accuracies of 95.21%, 94.17%, and 94.06%, respectively.
Authors: Justin Cui, Ruochen Wang, Yuanhao Xiong, Cho-Jui Hsieh
Abstract: Dataset Condensation has emerged as a technique for compressing large datasets into smaller synthetic counterparts, facilitating downstream training tasks. In this paper, we study the impact of bias inside the original dataset on the performance of dataset condensation. With a comprehensive empirical evaluation on canonical datasets with color, corruption and background biases, we found that color and background biases in the original dataset will be amplified through the condensation process, resulting in a notable decline in the performance of models trained on the condensed dataset, while corruption bias is suppressed through the condensation process. To reduce bias amplification in dataset condensation, we introduce a simple yet highly effective approach based on a sample reweighting scheme utilizing kernel density estimation. Empirical results on multiple real-world and synthetic datasets demonstrate the effectiveness of the proposed method. Notably, on CMNIST with 5% bias-conflict ratio and IPC 50, our method achieves 91.5% test accuracy compared to 23.8% from vanilla DM, boosting the performance by 67.7%, whereas applying state-of-the-art debiasing method on the same dataset only achieves 53.7% accuracy. Our findings highlight the importance of addressing biases in dataset condensation and provide a promising avenue to address bias amplification in the process.
Authors: Kai Liu, Haotong Qin, Yong Guo, Xin Yuan, Linghe Kong, Guihai Chen, Yulun Zhang
Abstract: Low-bit quantization has become widespread for compressing image super-resolution (SR) models for edge deployment, which allows advanced SR models to enjoy compact low-bit parameters and efficient integer/bitwise constructions for storage compression and inference acceleration, respectively. However, it is notorious that low-bit quantization degrades the accuracy of SR models compared to their full-precision (FP) counterparts. Despite several efforts to alleviate the degradation, the transformer-based SR model still suffers severe degradation due to its distinctive activation distribution. In this work, we present a dual-stage low-bit post-training quantization (PTQ) method for image super-resolution, namely 2DQuant, which achieves efficient and accurate SR under low-bit quantization. The proposed method first investigates the weight and activation and finds that the distribution is characterized by coexisting symmetry and asymmetry, long tails. Specifically, we propose Distribution-Oriented Bound Initialization (DOBI), using different searching strategies to search a coarse bound for quantizers. To obtain refined quantizer parameters, we further propose Distillation Quantization Calibration (DQC), which employs a distillation approach to make the quantized model learn from its FP counterpart. Through extensive experiments on different bits and scaling factors, the performance of DOBI can reach the state-of-the-art (SOTA) while after stage two, our method surpasses existing PTQ in both metrics and visual effects. 2DQuant gains an increase in PSNR as high as 4.52dB on Set5 (x2) compared with SOTA when quantized to 2-bit and enjoys a 3.60x compression ratio and 5.08x speedup ratio. The code and models will be available at https://github.com/Kai-Liu001/2DQuant.
Authors: Geongyu Lee, Joonho Lee, Tae-Yeong Kwak, Sun Woo Kim, Youngmee Kwon, Chungyeul Kim, Hyeyoon Chang
Abstract: Accurate prediction of the likelihood of recurrence is important in the selection of postoperative treatment for patients with early-stage breast cancer. In this study, we investigated whether deep learning algorithms can predict patients' risk of recurrence by analyzing the pathology images of their cancer histology. A total of 125 hematoxylin and eosin stained breast cancer whole slide images labeled with the risk prediction via genomics assays were used, and we obtained sensitivity of 0.857, 0.746, and 0.529 for predicting low, intermediate, and high risk, and specificity of 0.816, 0.803, and 0.972. When compared to the expert pathologist's regional histology grade information, a Pearson's correlation coefficient of 0.61 was obtained. When we checked the model learned through these studies through the class activation map, we found that it actually considered tubule formation and mitotic rate when predicting different risk groups.
Authors: Shester Gueuwou, Xiaodan Du, Greg Shakhnarovich, Karen Livescu
Abstract: A persistent challenge in sign language video processing, including the task of sign language to written language translation, is how we learn representations of sign language in an effective and efficient way that can preserve the important attributes of these languages, while remaining invariant to irrelevant visual differences. Informed by the nature and linguistics of signed languages, our proposed method focuses on just the most relevant parts in a signing video: the face, hands and body posture of the signer. However, instead of using pose estimation coordinates from off-the-shelf pose tracking models, which have inconsistent performance for hands and faces, we propose to learn the complex handshapes and rich facial expressions of sign languages in a self-supervised fashion. Our approach is based on learning from individual frames (rather than video sequences) and is therefore much more efficient than prior work on sign language pre-training. Compared to a recent model that established a new state of the art in sign language translation on the How2Sign dataset, our approach yields similar translation performance, using less than 3\% of the compute.
Authors: Elizabeth Newman, Katherine Keegan
Abstract: Recent advances in {matrix-mimetic} tensor frameworks have made it possible to preserve linear algebraic properties for multilinear data analysis and, as a result, to obtain optimal representations of multiway data. Matrix mimeticity arises from interpreting tensors as operators that can be multiplied, factorized, and analyzed analogous to matrices. Underlying the tensor operation is an algebraic framework parameterized by an invertible linear transformation. The choice of linear mapping is crucial to representation quality and, in practice, is made heuristically based on expected correlations in the data. However, in many cases, these correlations are unknown and common heuristics lead to suboptimal performance. In this work, we simultaneously learn optimal linear mappings and corresponding tensor representations without relying on prior knowledge of the data. Our new framework explicitly captures the coupling between the transformation and representation using variable projection. We preserve the invertibility of the linear mapping by learning orthogonal transformations with Riemannian optimization. We provide original theory of uniqueness of the transformation and convergence analysis of our variable-projection-based algorithm. We demonstrate the generality of our framework through numerical experiments on a wide range of applications, including financial index tracking, image compression, and reduced order modeling. We have published all the code related to this work at https://github.com/elizabethnewman/star-M-opt.
Authors: Yichi Zhang, Yao Huang, Yitong Sun, Chang Liu, Zhe Zhao, Zhengwei Fang, Yifan Wang, Huanran Chen, Xiao Yang, Xingxing Wei, Hang Su, Yinpeng Dong, Jun Zhu
Abstract: Despite the superior capabilities of Multimodal Large Language Models (MLLMs) across diverse tasks, they still face significant trustworthiness challenges. Yet, current literature on the assessment of trustworthy MLLMs remains limited, lacking a holistic evaluation to offer thorough insights into future improvements. In this work, we establish MultiTrust, the first comprehensive and unified benchmark on the trustworthiness of MLLMs across five primary aspects: truthfulness, safety, robustness, fairness, and privacy. Our benchmark employs a rigorous evaluation strategy that addresses both multimodal risks and cross-modal impacts, encompassing 32 diverse tasks with self-curated datasets. Extensive experiments with 21 modern MLLMs reveal some previously unexplored trustworthiness issues and risks, highlighting the complexities introduced by the multimodality and underscoring the necessity for advanced methodologies to enhance their reliability. For instance, typical proprietary models still struggle with the perception of visually confusing images and are vulnerable to multimodal jailbreaking and adversarial attacks; MLLMs are more inclined to disclose privacy in text and reveal ideological and cultural biases even when paired with irrelevant images in inference, indicating that the multimodality amplifies the internal risks from base LLMs. Additionally, we release a scalable toolbox for standardized trustworthiness research, aiming to facilitate future advancements in this important field. Code and resources are publicly available at: https://multi-trust.github.io/.
Authors: Gan Gao, Andrew H. Song, Fiona Wang, David Brenes, Rui Wang, Sarah S. L. Chow, Kevin W. Bishop, Lawrence D. True, Faisal Mahmood, Jonathan T. C. Liu
Abstract: Accurate patient diagnoses based on human tissue biopsies are hindered by current clinical practice, where pathologists assess only a limited number of thin 2D tissue slices sectioned from 3D volumetric tissue. Recent advances in non-destructive 3D pathology, such as open-top light-sheet microscopy, enable comprehensive imaging of spatially heterogeneous tissue morphologies, offering the feasibility to improve diagnostic determinations. A potential early route towards clinical adoption for 3D pathology is to rely on pathologists for final diagnosis based on viewing familiar 2D H&E-like image sections from the 3D datasets. However, manual examination of the massive 3D pathology datasets is infeasible. To address this, we present CARP3D, a deep learning triage approach that automatically identifies the highest-risk 2D slices within 3D volumetric biopsy, enabling time-efficient review by pathologists. For a given slice in the biopsy, we estimate its risk by performing attention-based aggregation of 2D patches within each slice, followed by pooling of the neighboring slices to compute a context-aware 2.5D risk score. For prostate cancer risk stratification, CARP3D achieves an area under the curve (AUC) of 90.4% for triaging slices, outperforming methods relying on independent analysis of 2D sections (AUC=81.3%). These results suggest that integrating additional depth context enhances the model's discriminative capabilities. In conclusion, CARP3D has the potential to improve pathologist diagnosis via accurate triage of high-risk slices within large-volume 3D pathology datasets.
Authors: Jinyuan Li, Ziyan Li, Han Li, Jianfei Yu, Rui Xia, Di Sun, Gang Pan
Abstract: Grounded Multimodal Named Entity Recognition (GMNER) task aims to identify named entities, entity types and their corresponding visual regions. GMNER task exhibits two challenging attributes: 1) The tenuous correlation between images and text on social media contributes to a notable proportion of named entities being ungroundable. 2) There exists a distinction between coarse-grained noun phrases used in similar tasks (e.g., phrase localization) and fine-grained named entities. In this paper, we propose RiVEG, a unified framework that reformulates GMNER into a joint MNER-VE-VG task by leveraging large language models (LLMs) as connecting bridges. This reformulation brings two benefits: 1) It enables us to optimize the MNER module for optimal MNER performance and eliminates the need to pre-extract region features using object detection methods, thus naturally addressing the two major limitations of existing GMNER methods. 2) The introduction of Entity Expansion Expression module and Visual Entailment (VE) module unifies Visual Grounding (VG) and Entity Grounding (EG). This endows the proposed framework with unlimited data and model scalability. Furthermore, to address the potential ambiguity stemming from the coarse-grained bounding box output in GMNER, we further construct the new Segmented Multimodal Named Entity Recognition (SMNER) task and corresponding Twitter-SMNER dataset aimed at generating fine-grained segmentation masks, and experimentally demonstrate the feasibility and effectiveness of using box prompt-based Segment Anything Model (SAM) to empower any GMNER model with the ability to accomplish the SMNER task. Extensive experiments demonstrate that RiVEG significantly outperforms SoTA methods on four datasets across the MNER, GMNER, and SMNER tasks.
Authors: Bo Zhou, Chuanzhao Lu, Yan Pan, Fu Chen
Abstract: Autonomous exploration in complex and cluttered environments is essential for various applications. However, there are many challenges due to the lack of global heuristic information. Existing exploration methods suffer from the repeated paths and considerable computational resource requirement in large-scale environments. To address the above issues, this letter proposes an efficient exploration planner that reduces repeated paths in complex environments, hence it is called "Only Travelling Once Planner". OTO Planner includes fast frontier updating, viewpoint evaluation and viewpoint refinement. A selective frontier updating mechanism is designed, saving a large amount of computational resources. In addition, a novel viewpoint evaluation system is devised to reduce the repeated paths utilizing the enclosed sub-region detection. Besides, a viewpoint refinement approach is raised to concentrate the redundant viewpoints, leading to smoother paths. We conduct extensive simulation and real-world experiments to validate the proposed method. Compared to the state-of-the-art approach, the proposed method reduces the exploration time and movement distance by 10%-20% and improves the speed of frontier detection by 6-9 times.
Authors: Adri\`a Molina, Oriol Ramos Terrades, Josep Llad\'os
Abstract: This paper introduces Fetch-A-Set (FAS), a comprehensive benchmark tailored for legislative historical document analysis systems, addressing the challenges of large-scale document retrieval in historical contexts. The benchmark comprises a vast repository of documents dating back to the XVII century, serving both as a training resource and an evaluation benchmark for retrieval systems. It fills a critical gap in the literature by focusing on complex extractive tasks within the domain of cultural heritage. The proposed benchmark tackles the multifaceted problem of historical document analysis, including text-to-image retrieval for queries and image-to-text topic extraction from document fragments, all while accommodating varying levels of document legibility. This benchmark aims to spur advancements in the field by providing baselines and data for the development and evaluation of robust historical document retrieval systems, particularly in scenarios characterized by wide historical spectrum.
Authors: Delfina Sol Martinez Pandiani, Erik Tjong Kim Sang, Davide Ceolin
Abstract: Internet memes, channels for humor, social commentary, and cultural expression, are increasingly used to spread toxic messages. Studies on the computational analyses of toxic memes have significantly grown over the past five years, and the only three surveys on computational toxic meme analysis cover only work published until 2022, leading to inconsistent terminology and unexplored trends. Our work fills this gap by surveying content-based computational perspectives on toxic memes, and reviewing key developments until early 2024. Employing the PRISMA methodology, we systematically extend the previously considered papers, achieving a threefold result. First, we survey 119 new papers, analyzing 158 computational works focused on content-based toxic meme analysis. We identify over 30 datasets used in toxic meme analysis and examine their labeling systems. Second, after observing the existence of unclear definitions of meme toxicity in computational works, we introduce a new taxonomy for categorizing meme toxicity types. We also note an expansion in computational tasks beyond the simple binary classification of memes as toxic or non-toxic, indicating a shift towards achieving a nuanced comprehension of toxicity. Third, we identify three content-based dimensions of meme toxicity under automatic study: target, intent, and conveyance tactics. We develop a framework illustrating the relationships between these dimensions and meme toxicities. The survey analyzes key challenges and recent trends, such as enhanced cross-modal reasoning, integrating expert and cultural knowledge, the demand for automatic toxicity explanations, and handling meme toxicity in low-resource languages. Also, it notes the rising use of Large Language Models (LLMs) and generative AI for detecting and generating toxic memes. Finally, it proposes pathways for advancing toxic meme detection and interpretation.
Authors: Abdurrahim Yilmaz, Sirin Pekcan Yasar, Gulsum Gencoglan, Burak Temelkuran
Abstract: Skin lesion datasets provide essential information for understanding various skin conditions and developing effective diagnostic tools. They aid the artificial intelligence-based early detection of skin cancer, facilitate treatment planning, and contribute to medical education and research. Published large datasets have partially coverage the subclassifications of the skin lesions. This limitation highlights the need for more expansive and varied datasets to reduce false predictions and help improve the failure analysis for skin lesions. This study presents a diverse dataset comprising 12,345 dermatoscopic images with 38 subclasses of skin lesions collected in Turkiye which comprises different skin types in the transition zone between Europe and Asia. Each subgroup contains high-resolution photos and expert annotations, providing a strong and reliable basis for future research. The detailed analysis of each subgroup provided in this study facilitates targeted research endeavors and enhances the depth of understanding regarding the skin lesions. This dataset distinguishes itself through a diverse structure with 5 super classes, 15 main classes, 38 subclasses and its 12,345 high-resolution dermatoscopic images.
Authors: Christopher D. Hsu, Pratik Chaudhari
Abstract: We study pursuit-evasion games in highly occluded urban environments, e.g. tall buildings in a city, where a scout (quadrotor) tracks multiple dynamic targets on the ground. We show that we can build a neural radiance field (NeRF) representation of the city -- online -- using RGB and depth images from different vantage points. This representation is used to calculate the information gain to both explore unknown parts of the city and track the targets -- thereby giving a completely first-principles approach to actively tracking dynamic targets. We demonstrate, using a custom-built simulator using Open Street Maps data of Philadelphia and New York City, that we can explore and locate 20 stationary targets within 300 steps. This is slower than a greedy baseline which which does not use active perception. But for dynamic targets that actively hide behind occlusions, we show that our approach maintains, at worst, a tracking error of 200m; the greedy baseline can have a tracking error as large as 600m. We observe a number of interesting properties in the scout's policies, e.g., it switches its attention to track a different target periodically, as the quality of the NeRF representation improves over time, the scout also becomes better in terms of target tracking.
Authors: Mason Wang, Ryosuke Sawata, Samuel Clarke, Ruohan Gao, Shangzhe Wu, Jiajun Wu
Abstract: Recent years have seen immense progress in 3D computer vision and computer graphics, with emerging tools that can virtualize real-world 3D environments for numerous Mixed Reality (XR) applications. However, alongside immersive visual experiences, immersive auditory experiences are equally vital to our holistic perception of an environment. In this paper, we aim to reconstruct the spatial acoustic characteristics of an arbitrary environment given only a sparse set of (roughly 12) room impulse response (RIR) recordings and a planar reconstruction of the scene, a setup that is easily achievable by ordinary users. To this end, we introduce DiffRIR, a differentiable RIR rendering framework with interpretable parametric models of salient acoustic features of the scene, including sound source directivity and surface reflectivity. This allows us to synthesize novel auditory experiences through the space with any source audio. To evaluate our method, we collect a dataset of RIR recordings and music in four diverse, real environments. We show that our model outperforms state-ofthe-art baselines on rendering monaural and binaural RIRs and music at unseen locations, and learns physically interpretable parameters characterizing acoustic properties of the sound source and surfaces in the scene.
Authors: Wenxiao Wang, Weiming Zhuang, Lingjuan Lyu
Abstract: The advancement of deep learning technologies is bringing new models every day, motivating the study of scalable model selection. An ideal model selection scheme should minimally support two operations efficiently over a large pool of candidate models: update, which involves either adding a new candidate model or removing an existing candidate model, and selection, which involves locating highly performing models for a given task. However, previous solutions to model selection require high computational complexity for at least one of these two operations. In this work, we target fundamentally (more) scalable model selection that supports asymptotically fast update and asymptotically fast selection at the same time. Firstly, we define isolated model embedding, a family of model selection schemes supporting asymptotically fast update and selection: With respect to the number of candidate models $m$, the update complexity is O(1) and the selection consists of a single sweep over $m$ vectors in addition to O(1) model operations. Isolated model embedding also implies several desirable properties for applications. Secondly, we present Standardized Embedder, an empirical realization of isolated model embedding. We assess its effectiveness by using it to select representations from a pool of 100 pre-trained vision models for classification tasks and measuring the performance gaps between the selected models and the best candidates with a linear probing protocol. Experiments suggest our realization is effective in selecting models with competitive performances and highlight isolated model embedding as a promising direction towards model selection that is fundamentally (more) scalable.
Authors: Tolga Bakirman, Elif Sertel
Abstract: Airplane detection from satellite imagery is a challenging task due to the complex backgrounds in the images and differences in data acquisition conditions caused by the sensor geometry and atmospheric effects. Deep learning methods provide reliable and accurate solutions for automatic detection of airplanes; however, huge amount of training data is required to obtain promising results. In this study, we create a novel airplane detection dataset called High Resolution Planes (HRPlanes) by using images from Google Earth (GE) and labeling the bounding box of each plane on the images. HRPlanes include GE images of several different airports across the world to represent a variety of landscape, seasonal and satellite geometry conditions obtained from different satellites. We evaluated our dataset with two widely used object detection methods namely YOLOv4 and Faster R-CNN. Our preliminary results show that the proposed dataset can be a valuable data source and benchmark data set for future applications. Moreover, proposed architectures and results of this study could be used for transfer learning of different datasets and models for airplane detection.
Authors: Ye Zhu, Yu Wu, Nicu Sebe, Yan Yan
Abstract: We are perceiving and communicating with the world in a multisensory manner, where different information sources are sophisticatedly processed and interpreted by separate parts of the human brain to constitute a complex, yet harmonious and unified sensing system. To endow the machines with true intelligence, multimodal machine learning that incorporates data from various sources has become an increasingly popular research area with emerging technical advances in recent years. In this paper, we present a survey on multimodal machine learning from a novel perspective considering not only the purely technical aspects but also the intrinsic nature of different data modalities. We analyze the commonness and uniqueness of each data format mainly ranging from vision, audio, text, and motions, and then present the methodological advancements categorized by the combination of data modalities, such as Vision+Text, with slightly inclined emphasis on the visual data. We investigate the existing literature on multimodal learning from both the representation learning and downstream application levels, and provide an additional comparison in the light of their technical connections with the data nature, e.g., the semantic consistency between image objects and textual descriptions, and the rhythm correspondence between video dance moves and musical beats. We hope that the exploitation of the alignment as well as the existing gap between the intrinsic nature of data modality and the technical designs, will benefit future research studies to better address a specific challenge related to the concrete multimodal task, prompting a unified multimodal machine learning framework closer to a real human intelligence system.
Authors: Yining Shi, Kun Jiang, Ke Wang, Jiusi Li, Yunlong Wang, Mengmeng Yang, Diange Yang
Abstract: Predicting the future occupancy states of the surrounding environment is a vital task for autonomous driving. However, current best-performing single-modality methods or multi-modality fusion perception methods are only able to predict uniform snapshots of future occupancy states and require strictly synchronized sensory data for sensor fusion. We propose a novel framework, StreamingFlow, to lift these strong limitations. StreamingFlow is a novel BEV occupancy predictor that ingests asynchronous multi-sensor data streams for fusion and performs streaming forecasting of the future occupancy map at any future timestamps. By integrating neural ordinary differential equations (N-ODE) into recurrent neural networks, StreamingFlow learns derivatives of BEV features over temporal horizons, updates the implicit sensor's BEV features as part of the fusion process, and propagates BEV states to the desired future time point. It shows good zero-shot generalization ability of prediction, reflected in the interpolation of the observed prediction time horizon and the reasonable inference of the unseen farther future period. Extensive experiments on two large-scale datasets, nuScenes and Lyft L5, demonstrate that StreamingFlow significantly outperforms previous vision-based, LiDAR-based methods, and shows superior performance compared to state-of-the-art fusion-based methods.
Authors: Wanqing Cui, Xin Hong, Yanyan Lan, Liang Pang, Jiafeng Guo, Xueqi Cheng
Abstract: Humans can naturally reason from superficial state differences (e.g. ground wetness) to transformations descriptions (e.g. raining) according to their life experience. In this paper, we propose a new visual reasoning task to test this transformation reasoning ability in real-world scenarios, called \textbf{V}isual \textbf{T}ransformation \textbf{T}elling (VTT). Given a series of states (i.e. images), VTT requires to describe the transformation occurring between every two adjacent states. Different from existing visual reasoning tasks that focus on surface state reasoning, the advantage of VTT is that it captures the underlying causes, e.g. actions or events, behind the differences among states. We collect a novel dataset to support the study of transformation reasoning from two existing instructional video datasets, CrossTask and COIN, comprising 13,547 samples. Each sample involves the key state images along with their transformation descriptions. Our dataset covers diverse real-world activities, providing a rich resource for training and evaluation. To construct an initial benchmark for VTT, we test several models, including traditional visual storytelling methods (CST, GLACNet, Densecap) and advanced multimodal large language models (LLaVA v1.5-7B, Qwen-VL-chat, Gemini Pro Vision, GPT-4o, and GPT-4). Experimental results reveal that even state-of-the-art models still face challenges in VTT, highlighting substantial areas for improvement.
Authors: Avinash Madasu, Vasudev Lal
Abstract: Video retrieval (VR) involves retrieving the ground truth video from the video database given a text caption or vice-versa. The two important components of compositionality: objects & attributes and actions are joined using correct syntax to form a proper text query. These components (objects & attributes, actions and syntax) each play an important role to help distinguish among videos and retrieve the correct ground truth video. However, it is unclear what is the effect of these components on the video retrieval performance. We therefore, conduct a systematic study to evaluate the compositional and syntactic understanding of video retrieval models on standard benchmarks such as MSRVTT, MSVD and DIDEMO. The study is performed on two categories of video retrieval models: (i) which are pre-trained on video-text pairs and fine-tuned on downstream video retrieval datasets (Eg. Frozen-in-Time, Violet, MCQ etc.) (ii) which adapt pre-trained image-text representations like CLIP for video retrieval (Eg. CLIP4Clip, XCLIP, CLIP2Video etc.). Our experiments reveal that actions and syntax play a minor role compared to objects & attributes in video understanding. Moreover, video retrieval models that use pre-trained image-text representations (CLIP) have better syntactic and compositional understanding as compared to models pre-trained on video-text data. The code is available at https://github.com/IntelLabs/multimodal_cognitive_ai/tree/main/ICSVR
URLs: https://github.com/IntelLabs/multimodal_cognitive_ai/tree/main/ICSVR
Authors: He Wang, Yunfeng Diao
Abstract: Classifiers based on deep neural networks are susceptible to adversarial attack, where the widely existing vulnerability has invoked the research in defending them from potential threats. Given a vulnerable classifier, existing defense methods are mostly white-box and often require re-training the victim under modified loss functions/training regimes. While the model/data/training specifics of the victim are usually unavailable to the user, re-training is unappealing, if not impossible for reasons such as limited computational resources. To this end, we propose a new post-train black-box defense framework. It can turn any pre-trained classifier into a resilient one with little knowledge of the model specifics. This is achieved by new joint Bayesian treatments on the clean data, the adversarial examples and the classifier, for maximizing their joint probability. It is further equipped with a new post-train strategy which keeps the victim intact, avoiding re-training. We name our framework Bayesian Boundary Correction (BBC). BBC is a general and flexible framework that can easily adapt to different data types. We instantiate BBC for image classification and skeleton-based human activity recognition, for both static and dynamic data. Exhaustive evaluation shows that BBC has superior robustness and can enhance robustness without severely hurting the clean accuracy, compared with existing defense methods.
Authors: Mohammad Adiban, Kalin Stefanov, Sabato Marco Siniscalchi, Giampiero Salvi
Abstract: We address the video prediction task by putting forth a novel model that combines (i) our recently proposed hierarchical residual vector quantized variational autoencoder (HR-VQVAE), and (ii) a novel spatiotemporal PixelCNN (ST-PixelCNN). We refer to this approach as a sequential hierarchical residual learning vector quantized variational autoencoder (S-HR-VQVAE). By leveraging the intrinsic capabilities of HR-VQVAE at modeling still images with a parsimonious representation, combined with the ST-PixelCNN's ability at handling spatiotemporal information, S-HR-VQVAE can better deal with chief challenges in video prediction. These include learning spatiotemporal information, handling high dimensional data, combating blurry prediction, and implicit modeling of physical characteristics. Extensive experimental results on the KTH Human Action and Moving-MNIST tasks demonstrate that our model compares favorably against top video prediction techniques both in quantitative and qualitative evaluations despite a much smaller model size. Finally, we boost S-HR-VQVAE by proposing a novel training method to jointly estimate the HR-VQVAE and ST-PixelCNN parameters.
Authors: Jialun Pei, Tao Jiang, He Tang, Nian Liu, Yueming Jin, Deng-Ping Fan, Pheng-Ann Heng
Abstract: We propose a novel approach for RGB-D salient instance segmentation using a dual-branch cross-modal feature calibration architecture called CalibNet. Our method simultaneously calibrates depth and RGB features in the kernel and mask branches to generate instance-aware kernels and mask features. CalibNet consists of three simple modules, a dynamic interactive kernel (DIK) and a weight-sharing fusion (WSF), which work together to generate effective instance-aware kernels and integrate cross-modal features. To improve the quality of depth features, we incorporate a depth similarity assessment (DSA) module prior to DIK and WSF. In addition, we further contribute a new DSIS dataset, which contains 1,940 images with elaborate instance-level annotations. Extensive experiments on three challenging benchmarks show that CalibNet yields a promising result, i.e., 58.0% AP with 320*480 input size on the COME15K-N test set, which significantly surpasses the alternative frameworks. Our code and dataset are available at: https://github.com/PJLallen/CalibNet.
Authors: Hu Zhang, Yanchen Li, Luziwei Leng, Kaiwei Che, Qian Liu, Qinghai Guo, Jianxing Liao, Ran Cheng
Abstract: Event-based sensors, distinguished by their high temporal resolution of 1 $\mathrm{\mu}\text{s}$ and a dynamic range of 120 $\text{dB}$, stand out as ideal tools for deployment in fast-paced settings like vehicles and drones. Traditional object detection techniques that utilize Artificial Neural Networks (ANNs) face challenges due to the sparse and asynchronous nature of the events these sensors capture. In contrast, Spiking Neural Networks (SNNs) offer a promising alternative, providing a temporal representation that is inherently aligned with event-based data. This paper explores the unique membrane potential dynamics of SNNs and their ability to modulate sparse events. We introduce an innovative spike-triggered adaptive threshold mechanism designed for stable training. Building on these insights, we present a specialized spiking feature pyramid network (SpikeFPN) optimized for automotive event-based object detection. Comprehensive evaluations demonstrate that SpikeFPN surpasses both traditional SNNs and advanced ANNs enhanced with attention mechanisms. Evidently, SpikeFPN achieves a mean Average Precision (mAP) of 0.477 on the GEN1 Automotive Detection (GAD) benchmark dataset, marking significant increases over the selected SNN baselines. Moreover, the efficient design of SpikeFPN ensures robust performance while optimizing computational resources, attributed to its innate sparse computation capabilities. Source codes are publicly accessible at https://github.com/EMI-Group/spikefpn.
Authors: Jianhua Qiua, Weihua Liu, Chaochao Lin, Jiaojiao Li, Haoping Yu, Said Boumaraf
Abstract: Face parsing infers a pixel-wise label map for each semantic facial component. Previous methods generally work well for uncovered faces, however, they overlook facial occlusion and ignore some contextual areas outside a single face, especially when facial occlusion has become a common situation during the COVID-19 epidemic. Inspired by the lighting phenomena in everyday life, where illumination from four distinct lamps provides a more uniform distribution than a single central light source, we propose a novel homogeneous tanh-transform for image preprocessing, which is made up of four tanh-transforms. These transforms fuse the central vision and the peripheral vision together. Our proposed method addresses the dilemma of face parsing under occlusion and compresses more information from the surrounding context. Based on homogeneous tanh-transforms, we propose an occlusion-aware convolutional neural network for occluded face parsing. It combines information in both Tanh-polar space and Tanh-Cartesian space, capable of enhancing receptive fields. Furthermore, we introduce an occlusion-aware loss to focus on the boundaries of occluded regions. The network is simple, flexible, and can be trained end-to-end. To facilitate future research of occluded face parsing, we also contribute a new cleaned face parsing dataset. This dataset is manually purified from several academic or industrial datasets, including CelebAMask-HQ, Short-video Face Parsing, and the Helen dataset, and will be made public. Experiments demonstrate that our method surpasses state-of-the-art methods in face parsing under occlusion.
Authors: Florent Forest, Hugo Porta, Devis Tuia, Olga Fink
Abstract: Monitoring surface cracks in infrastructure is crucial for structural health monitoring. Automatic visual inspection offers an effective solution, especially in hard-to-reach areas. Machine learning approaches have proven their effectiveness but typically require large annotated datasets for supervised training. Once a crack is detected, monitoring its severity often demands precise segmentation of the damage. However, pixel-level annotation of images for segmentation is labor-intensive. To mitigate this cost, one can leverage explainable artificial intelligence (XAI) to derive segmentations from the explanations of a classifier, requiring only weak image-level supervision. This paper proposes applying this methodology to segment and monitor surface cracks. We evaluate the performance of various XAI methods and examine how this approach facilitates severity quantification and growth monitoring. Results reveal that while the resulting segmentation masks may exhibit lower quality than those produced by supervised methods, they remain meaningful and enable severity monitoring, thus reducing substantial labeling costs.
Authors: Haoxiang Wang, Pavan Kumar Anasosalu Vasu, Fartash Faghri, Raviteja Vemulapalli, Mehrdad Farajtabar, Sachin Mehta, Mohammad Rastegari, Oncel Tuzel, Hadi Pouransari
Abstract: The landscape of publicly available vision foundation models (VFMs), such as CLIP and Segment Anything Model (SAM), is expanding rapidly. VFMs are endowed with distinct capabilities stemming from their pre-training objectives. For instance, CLIP excels in semantic understanding, while SAM specializes in spatial understanding for segmentation. In this work, we introduce a simple recipe to efficiently merge VFMs into a unified model that absorbs their expertise. Our method integrates techniques of multi-task learning, continual learning, and distillation. Further, it demands significantly less computational cost compared to traditional multi-task training from scratch, and it only needs a small fraction of the pre-training datasets that were initially used to train individual models. By applying our method to SAM and CLIP, we obtain SAM-CLIP: a unified model that combines the capabilities of SAM and CLIP into a single vision transformer. Compared with deploying SAM and CLIP independently, our merged model, SAM-CLIP, reduces storage and compute costs for inference, making it well-suited for edge device applications. We show that SAM-CLIP not only retains the foundational strengths of SAM and CLIP, but also introduces synergistic functionalities, notably in zero-shot semantic segmentation, where SAM-CLIP establishes new state-of-the-art results on 5 benchmarks. It outperforms previous models that are specifically designed for this task by a large margin, including +6.8% and +5.9% mean IoU improvement on Pascal-VOC and COCO-Stuff datasets, respectively.
Authors: Jan Zah\'alka
Abstract: Adversarial attacks are an important security concern for computer vision (CV). As CV models are becoming increasingly valuable assets in applied practice, disrupting them is emerging as a form of economic sabotage. This paper opens up the exploration of damaging adversarial attacks (DAAs) that seek to damage target CV models. DAAs are formalized by defining the threat model, the cost function DAAs maximize, and setting three requirements for success: potency, stealth, and customizability. As a pioneer DAA, this paper proposes Trainwreck, a train-time attack that conflates the data of similar classes in the training data using stealthy ($\epsilon \leq 8/255$) class-pair universal perturbations obtained from a surrogate model. Trainwreck is a black-box, transferable attack: it requires no knowledge of the target architecture, and a single poisoned dataset degrades the performance of any model trained on it. The experimental evaluation on CIFAR-10 and CIFAR-100 and various model architectures (EfficientNetV2, ResNeXt-101, and a finetuned ViT-L-16) demonstrates Trainwreck's efficiency. Trainwreck achieves similar or better potency compared to the data poisoning state of the art and is fully customizable by the poison rate parameter. Finally, data redundancy with hashing is identified as a reliable defense against Trainwreck or similar DAAs. The code is available at https://github.com/JanZahalka/trainwreck.
Authors: Xiujun Li, Yujie Lu, Zhe Gan, Jianfeng Gao, William Yang Wang, Yejin Choi
Abstract: Recent multimodal large language models (MLLMs) have shown promising instruction following capabilities on vision-language tasks. In this work, we introduce VISUAL MODALITY INSTRUCTION (VIM), and investigate how well multimodal models can understand textual instructions provided in pixels, despite not being explicitly trained on such data during pretraining or fine-tuning. We adapt VIM to eight benchmarks, including OKVQA, MM-Vet, MathVista, MMMU, and probe diverse MLLMs in both the text-modality instruction (TEM) setting and VIM setting. Notably, we observe a significant performance disparity between the original TEM and VIM settings for open-source MLLMs, indicating that open-source MLLMs face greater challenges when text instruction is presented solely in image form. To address this issue, we train v-MLLM, a generalizable model that is capable to conduct robust instruction following in both text-modality and visual-modality instructions.
Authors: KL Navaneet, Kossar Pourahmadi Meibodi, Soroush Abbasi Koohpayegani, Hamed Pirsiavash
Abstract: 3D Gaussian Splatting (3DGS) is a new method for modeling and rendering 3D radiance fields that achieves much faster learning and rendering time compared to SOTA NeRF methods. However, it comes with the drawback of a much larger storage demand compared to NeRF methods since it needs to store the parameters for millions of 3D Gaussians. We notice that large groups of Gaussians share similar parameters and introduce a simple vector quantization method based on K-means algorithm to quantize the Gaussian parameters. Then, we store the small codebook along with the index of the code for each Gaussian. We compress the indices further by sorting them and using a method similar to run-length encoding. Moreover, we use a simple regularizer that encourages zero opacity (invisible Gaussians) to reduce the number of Gaussians, thereby compressing the model and speeding up the rendering. We do extensive experiments on standard benchmarks as well as an existing 3D dataset that is an order of magnitude larger than the standard benchmarks used in this field. We show that our simple yet effective method can reduce the storage costs for 3DGS by 40 to 50x and rendering time by 2 to 3x with a very small drop in the quality of rendered images.
Authors: Gege Gao, Weiyang Liu, Anpei Chen, Andreas Geiger, Bernhard Sch\"olkopf
Abstract: As pretrained text-to-image diffusion models become increasingly powerful, recent efforts have been made to distill knowledge from these text-to-image pretrained models for optimizing a text-guided 3D model. Most of the existing methods generate a holistic 3D model from a plain text input. This can be problematic when the text describes a complex scene with multiple objects, because the vectorized text embeddings are inherently unable to capture a complex description with multiple entities and relationships. Holistic 3D modeling of the entire scene further prevents accurate grounding of text entities and concepts. To address this limitation, we propose GraphDreamer, a novel framework to generate compositional 3D scenes from scene graphs, where objects are represented as nodes and their interactions as edges. By exploiting node and edge information in scene graphs, our method makes better use of the pretrained text-to-image diffusion model and is able to fully disentangle different objects without image-level supervision. To facilitate modeling of object-wise relationships, we use signed distance fields as representation and impose a constraint to avoid inter-penetration of objects. To avoid manual scene graph creation, we design a text prompt for ChatGPT to generate scene graphs based on text inputs. We conduct both qualitative and quantitative experiments to validate the effectiveness of GraphDreamer in generating high-fidelity compositional 3D scenes with disentangled object entities.
Authors: Ziyu Zhou, Haozhe Luo, Jiaxuan Pang, Xiaowei Ding, Michael Gotway, Jianming Liang
Abstract: Self-supervised learning (SSL) approaches have recently shown substantial success in learning visual representations from unannotated images. Compared with photographic images, medical images acquired with the same imaging protocol exhibit high consistency in anatomy. To exploit this anatomical consistency, this paper introduces a novel SSL approach, called PEAC (patch embedding of anatomical consistency), for medical image analysis. Specifically, in this paper, we propose to learn global and local consistencies via stable grid-based matching, transfer pre-trained PEAC models to diverse downstream tasks, and extensively demonstrate that (1) PEAC achieves significantly better performance than the existing state-of-the-art fully/self-supervised methods, and (2) PEAC captures the anatomical structure consistency across views of the same patient and across patients of different genders, weights, and healthy statuses, which enhances the interpretability of our method for medical image analysis.
Authors: Yifan Liu, Tiecheng Song, Chengye Xian, Ruiyuan Chen, Yi Zhao, Rui Li, Tan Guo
Abstract: Crater detection can provide valuable information for humans to explore the topography and understand the history of extraterrestrial planets. Due to the significantly varying scenario distributions, existing detection models trained on known labelled crater datasets are hardly effective when applied to new unlabelled planets. To address this issue, we propose a two-stage adaptive network (TAN) for semi-supervised cross-domain crater detection. Our network is built on the YOLOv5 detector, where a series of strategies are employed to enhance its cross-domain generalisation ability. In the first stage, we propose an attention-based scale-adaptive fusion (ASAF) strategy to handle objects with significant scale variances. Furthermore, we propose a smoothing hard example mining (SHEM) loss function to address the issue of overfitting on hard examples. In the second stage, we propose a sort-based pseudo-labelling fine-tuning (SPF) strategy for semi-supervised learning to mitigate the distributional differences between source and target domains. For both stages, we employ weak or strong image augmentation to suit different cross-domain tasks. Experimental results on benchmark datasets demonstrate that the proposed network can enhance domain adaptation ability for crater detection under varying scenario distributions.
Authors: Yi Chen, Yuying Ge, Yixiao Ge, Mingyu Ding, Bohao Li, Rui Wang, Ruifeng Xu, Ying Shan, Xihui Liu
Abstract: The pursuit of artificial general intelligence (AGI) has been accelerated by Multimodal Large Language Models (MLLMs), which exhibit superior reasoning, generalization capabilities, and proficiency in processing multimodal inputs. A crucial milestone in the evolution of AGI is the attainment of human-level planning, a fundamental ability for making informed decisions in complex environments, and solving a wide range of real-world problems. Despite the impressive advancements in MLLMs, a question remains: How far are current MLLMs from achieving human-level planning? To shed light on this question, we introduce EgoPlan-Bench, a comprehensive benchmark to evaluate the planning abilities of MLLMs in real-world scenarios from an egocentric perspective, mirroring human perception. EgoPlan-Bench emphasizes the evaluation of planning capabilities of MLLMs, featuring realistic tasks, diverse action plans, and intricate visual observations. Our rigorous evaluation of a wide range of MLLMs reveals that EgoPlan-Bench poses significant challenges, highlighting a substantial scope for improvement in MLLMs to achieve human-level task planning. To facilitate this advancement, we further present EgoPlan-IT, a specialized instruction-tuning dataset that effectively enhances model performance on EgoPlan-Bench. We have made all codes, data, and a maintained benchmark leaderboard available to advance future research.
Authors: Dawid Malarz, Weronika Smolak, Jacek Tabor, S{\l}awomir Tadeja, Przemys{\l}aw Spurek
Abstract: Neural Radiance Fields (NeRFs) have demonstrated the remarkable potential of neural networks to capture the intricacies of 3D objects. By encoding the shape and color information within neural network weights, NeRFs excel at producing strikingly sharp novel views of 3D objects. Recently, numerous generalizations of NeRFs utilizing generative models have emerged, expanding its versatility. In contrast, Gaussian Splatting (GS) offers a similar render quality with faster training and inference as it does not need neural networks to work. It encodes information about the 3D objects in the set of Gaussian distributions that can be rendered in 3D similarly to classical meshes. Unfortunately, GS are difficult to condition since they usually require circa hundred thousand Gaussian components. To mitigate the caveats of both models, we propose a hybrid model Viewing Direction Gaussian Splatting (VDGS) that uses GS representation of the 3D object's shape and NeRF-based encoding of color and opacity. Our model uses Gaussian distributions with trainable positions (i.e. means of Gaussian), shape (i.e. covariance of Gaussian), color and opacity, and a neural network that takes Gaussian parameters and viewing direction to produce changes in the said color and opacity. As a result, our model better describes shadows, light reflections, and the transparency of 3D objects without adding additional texture and light components.
Authors: Zeyu Li, Suncheng Xiang, Tong Yu, Jingsheng Gao, Jiacheng Ruan, Yanping Hu, Ting Liu, Yuzhuo Fu
Abstract: The recognition of underwater audio plays a significant role in identifying a vessel while it is in motion. Underwater target recognition tasks have a wide range of applications in areas such as marine environmental protection, detection of ship radiated noise, underwater noise control, and coastal vessel dispatch. The traditional UATR task involves training a network to extract features from audio data and predict the vessel type. The current UATR dataset exhibits shortcomings in both duration and sample quantity. In this paper, we propose Oceanship, a large-scale and diverse underwater audio dataset. This dataset comprises 15 categories, spans a total duration of 121 hours, and includes comprehensive annotation information such as coordinates, velocity, vessel types, and timestamps. We compiled the dataset by crawling and organizing original communication data from the Ocean Communication Network (ONC) database between 2021 and 2022. While audio retrieval tasks are well-established in general audio classification, they have not been explored in the context of underwater audio recognition. Leveraging the Oceanship dataset, we introduce a baseline model named Oceannet for underwater audio retrieval. This model achieves a recall at 1 (R@1) accuracy of 67.11% and a recall at 5 (R@5) accuracy of 99.13% on the Deepship dataset.
Authors: Sunyi Zheng, Xiaonan Cui, Yuxuan Sun, Jingxiong Li, Honglin Li, Yunlong Zhang, Pingyi Chen, Xueping Jing, Zhaoxiang Ye, Lin Yang
Abstract: Accurate image classification and retrieval are of importance for clinical diagnosis and treatment decision-making. The recent contrastive language-image pretraining (CLIP) model has shown remarkable proficiency in understanding natural images. Drawing inspiration from CLIP, PathCLIP is specifically designed for pathology image analysis, utilizing over 200,000 image and text pairs in training. While the performance the PathCLIP is impressive, its robustness under a wide range of image corruptions remains unknown. Therefore, we conduct an extensive evaluation to analyze the performance of PathCLIP on various corrupted images from the datasets of Osteosarcoma and WSSS4LUAD. In our experiments, we introduce seven corruption types including brightness, contrast, Gaussian blur, resolution, saturation, hue, and markup at four severity levels. Through experiments, we find that PathCLIP is relatively robustness to image corruptions and surpasses OpenAI-CLIP and PLIP in zero-shot classification. Among the seven corruptions, blur and resolution can cause server performance degradation of the PathCLIP. This indicates that ensuring the quality of images is crucial before conducting a clinical test. Additionally, we assess the robustness of PathCLIP in the task of image-image retrieval, revealing that PathCLIP performs less effectively than PLIP on Osteosarcoma but performs better on WSSS4LUAD under diverse corruptions. Overall, PathCLIP presents impressive zero-shot classification and retrieval performance for pathology images, but appropriate care needs to be taken when using it. We hope this study provides a qualitative impression of PathCLIP and helps understand its differences from other CLIP models.
Authors: Nathan Mankovich, Gustau Camps-Valls, Tolga Birdal
Abstract: Principal component analysis (PCA), along with its extensions to manifolds and outlier contaminated data, have been indispensable in computer vision and machine learning. In this work, we present a unifying formalism for PCA and its variants, and introduce a framework based on the flags of linear subspaces, ie a hierarchy of nested linear subspaces of increasing dimension, which not only allows for a common implementation but also yields novel variants, not explored previously. We begin by generalizing traditional PCA methods that either maximize variance or minimize reconstruction error. We expand these interpretations to develop a wide array of new dimensionality reduction algorithms by accounting for outliers and the data manifold. To devise a common computational approach, we recast robust and dual forms of PCA as optimization problems on flag manifolds. We then integrate tangent space approximations of principal geodesic analysis (tangent-PCA) into this flag-based framework, creating novel robust and dual geodesic PCA variations. The remarkable flexibility offered by the 'flagification' introduced here enables even more algorithmic variants identified by specific flag types. Last but not least, we propose an effective convergent solver for these flag-formulations employing the Stiefel manifold. Our empirical results on both real-world and synthetic scenarios, demonstrate the superiority of our novel algorithms, especially in terms of robustness to outliers on manifolds.
Authors: Tuan Anh Le, Pavel Sountsov, Matthew D. Hoffman, Ben Lee, Brian Patton, Rif A. Saurous
Abstract: How do we infer a 3D scene from a single image in the presence of corruptions like rain, snow or fog? Straightforward domain randomization relies on knowing the family of corruptions ahead of time. Here, we propose a Bayesian approach-dubbed robust inverse graphics (RIG)-that relies on a strong scene prior and an uninformative uniform corruption prior, making it applicable to a wide range of corruptions. Given a single image, RIG performs posterior inference jointly over the scene and the corruption. We demonstrate this idea by training a neural radiance field (NeRF) scene prior and using a secondary NeRF to represent the corruptions over which we place an uninformative prior. RIG, trained only on clean data, outperforms depth estimators and alternative NeRF approaches that perform point estimation instead of full inference. The results hold for a number of scene prior architectures based on normalizing flows and diffusion models. For the latter, we develop reconstruction-guidance with auxiliary latents (ReGAL)-a diffusion conditioning algorithm that is applicable in the presence of auxiliary latent variables such as the corruption. RIG demonstrates how scene priors can be used beyond generation tasks.
Authors: Xiaoyu Zhou, Xingjian Ran, Yajiao Xiong, Jinlin He, Zhiwei Lin, Yongtao Wang, Deqing Sun, Ming-Hsuan Yang
Abstract: We present GALA3D, generative 3D GAussians with LAyout-guided control, for effective compositional text-to-3D generation. We first utilize large language models (LLMs) to generate the initial layout and introduce a layout-guided 3D Gaussian representation for 3D content generation with adaptive geometric constraints. We then propose an instance-scene compositional optimization mechanism with conditioned diffusion to collaboratively generate realistic 3D scenes with consistent geometry, texture, scale, and accurate interactions among multiple objects while simultaneously adjusting the coarse layout priors extracted from the LLMs to align with the generated scene. Experiments show that GALA3D is a user-friendly, end-to-end framework for state-of-the-art scene-level 3D content generation and controllable editing while ensuring the high fidelity of object-level entities within the scene. The source codes and models will be available at gala3d.github.io.
Authors: Zhaoqing Wang, Xiaobo Xia, Ziye Chen, Xiao He, Yandong Guo, Mingming Gong, Tongliang Liu
Abstract: Current state-of-the-art open-vocabulary segmentation methods typically rely on image-mask-text triplet annotations for supervision. However, acquiring such detailed annotations is labour-intensive and poses scalability challenges in complex real-world scenarios. While existing weakly-supervised approaches leverage image-text pairs to reduce the expansive annotation cost, the lack of mask supervision makes it difficult for the model to locate multiple instances and accurately group pixels with similar semantics, significantly hampering versatility and performance. In this paper, we introduce Unpair-Seg, a novel weakly-supervised open-vocabulary segmentation framework that learns from unpaired image-mask and image-text pairs, which can be independently and efficiently collected. Unpair-Seg initially predicts a set of binary masks and generates pseudo labels by identifying confident pairs of masks and text entities. We then train a feature adapter to align region embeddings with text embeddings based on these pseudo labels, achieving open-vocabulary segmentation. However, the inherent noise in the mask-entity correspondence poses a challenge to obtaining reliable pairs. To address this, we employ a vision-language large model to re-caption the input images and extract precise entities, and we design a multi-scale matching strategy to reduce noisy mask-entity pairs. Our Unpair-Seg framework demonstrates impressive performance, achieving 14.6\% and 19.5\% mIoU on the ADE-847 and PASCAL Context-459 datasets, significantly narrowing the gap between fully-supervised and weakly-supervised methods.
Authors: Jiakai Sun, Han Jiao, Guangyuan Li, Zhanjie Zhang, Lei Zhao, Wei Xing
Abstract: Constructing photo-realistic Free-Viewpoint Videos (FVVs) of dynamic scenes from multi-view videos remains a challenging endeavor. Despite the remarkable advancements achieved by current neural rendering techniques, these methods generally require complete video sequences for offline training and are not capable of real-time rendering. To address these constraints, we introduce 3DGStream, a method designed for efficient FVV streaming of real-world dynamic scenes. Our method achieves fast on-the-fly per-frame reconstruction within 12 seconds and real-time rendering at 200 FPS. Specifically, we utilize 3D Gaussians (3DGs) to represent the scene. Instead of the na\"ive approach of directly optimizing 3DGs per-frame, we employ a compact Neural Transformation Cache (NTC) to model the translations and rotations of 3DGs, markedly reducing the training time and storage required for each FVV frame. Furthermore, we propose an adaptive 3DG addition strategy to handle emerging objects in dynamic scenes. Experiments demonstrate that 3DGStream achieves competitive performance in terms of rendering speed, image quality, training time, and model storage when compared with state-of-the-art methods.
Authors: Yao Jiang, Xinyu Yan, Ge-Peng Ji, Keren Fu, Meijun Sun, Huan Xiong, Deng-Ping Fan, Fahad Shahbaz Khan
Abstract: The advent of large vision-language models (LVLMs) represents a remarkable advance in the quest for artificial general intelligence. However, the model's effectiveness in both specialized and general tasks warrants further investigation. This paper endeavors to evaluate the competency of popular LVLMs in specialized and general tasks, respectively, aiming to offer a comprehensive understanding of these novel models. To gauge their effectiveness in specialized tasks, we employ six challenging tasks in three different application scenarios: natural, healthcare, and industrial. These six tasks include salient/camouflaged/transparent object detection, as well as polyp detection, skin lesion detection, and industrial anomaly detection. We examine the performance of three recent open-source LVLMs, including MiniGPT-v2, LLaVA-1.5, and Shikra, on both visual recognition and localization in these tasks. Moreover, we conduct empirical investigations utilizing the aforementioned LVLMs together with GPT-4V, assessing their multi-modal understanding capabilities in general tasks including object counting, absurd question answering, affordance reasoning, attribute recognition, and spatial relation reasoning. Our investigations reveal that these LVLMs demonstrate limited proficiency not only in specialized tasks but also in general tasks. We delve deep into this inadequacy and uncover several potential factors, including limited cognition in specialized tasks, object hallucination, text-to-image interference, and decreased robustness in complex problems. We hope that this study can provide useful insights for the future development of LVLMs, helping researchers improve LVLMs for both general and specialized applications.
Authors: Datao Tang, Xiangyong Cao, Xingsong Hou, Zhongyuan Jiang, Deyu Meng
Abstract: The emergence of generative models has revolutionized the field of remote sensing (RS) image generation. Despite generating high-quality images, existing methods are limited in relying mainly on text control conditions and thus don't always generate images accurately and stablely. In this paper, we propose CRS-Diff, a new RS generative foundation framework specifically tailored for RS image generation, leveraging the inherent advantages of diffusion models while integrating more advanced control mechanisms. Specifically, CRS-Diff can simultaneously support text-condition, metadata-condition, and image-condition control inputs, thus enabling more precise control to refine the generation process. To effectively integrate multiple condition control information, we introduce a new conditional control mechanism to achieve multi-scale feature fusion, thus enhancing the guiding effect of control conditions. To our knowledge, CRS-Diff is the first multiple-condition controllable generative RS foundation model. Experimental results in single-condition and multiple-condition cases have demonstrated the superior ability of our CRS-Diff to generate RS images both quantitatively and qualitatively compared with previous methods. Additionally, our CRS-Diff can serve as a data engine that generates high-quality training data for downstream tasks, e.g., road extraction. The code is available at https://github.com/Sonettoo/CRS-Diff.
Authors: Max Ku, Cong Wei, Weiming Ren, Harry Yang, Wenhu Chen
Abstract: In the dynamic field of digital content creation using generative models, state-of-the-art video editing models still do not offer the level of quality and control that users desire. Previous works on video editing either extended from image-based generative models in a zero-shot manner or necessitated extensive fine-tuning, which can hinder the production of fluid video edits. Furthermore, these methods frequently rely on textual input as the editing guidance, leading to ambiguities and limiting the types of edits they can perform. Recognizing these challenges, we introduce AnyV2V, a novel tuning-free paradigm designed to simplify video editing into two primary steps: (1) employing an off-the-shelf image editing model to modify the first frame, (2) utilizing an existing image-to-video generation model to generate the edited video through temporal feature injection. AnyV2V can leverage any existing image editing tools to support an extensive array of video editing tasks, including prompt-based editing, reference-based style transfer, subject-driven editing, and identity manipulation, which were unattainable by previous methods. AnyV2V can also support any video length. Our evaluation indicates that AnyV2V significantly outperforms other baseline methods in automatic and human evaluations by significant margin, maintaining visual consistency with the source video while achieving high-quality edits across all the editing tasks.
Authors: Mahdi Naseri, Jiayan Qiu, Zhou Wang
Abstract: In this paper, we investigate a novel artificial intelligence generation task, termed as generated contents enrichment (GCE). Different from conventional artificial intelligence contents generation task that enriches the given textual description implicitly with limited semantics for generating visually real content, our proposed GCE strives to perform content enrichment explicitly on both the visual and textual domain, from which the enriched contents are visually real, structurally reasonable, and semantically abundant. Towards to solve GCE, we propose a deep end-to-end method that explicitly explores the semantics and inter-semantic relationships during the enrichment. Specifically, we first model the input description as a semantic graph, wherein each node represents an object and each edge corresponds to the inter-object relationship. We then adopt Graph Convolutional Networks on top of the input scene description to predict the enriching objects and their relationships with the input objects. Finally, the enriched description is fed into an image synthesis model to carry out the visual contents generation. Our experiments conducted on the Visual Genome dataset exhibit promising and visually plausible results.
Authors: Kaiwen Jiang, Yang Fu, Mukund Varma T, Yash Belhe, Xiaolong Wang, Hao Su, Ravi Ramamoorthi
Abstract: Novel view synthesis from a sparse set of input images is a challenging problem of great practical interest, especially when camera poses are absent or inaccurate. Direct optimization of camera poses and usage of estimated depths in neural radiance field algorithms usually do not produce good results because of the coupling between poses and depths, and inaccuracies in monocular depth estimation. In this paper, we leverage the recent 3D Gaussian splatting method to develop a novel construct-and-optimize method for sparse view synthesis without camera poses. Specifically, we construct a solution progressively by using monocular depth and projecting pixels back into the 3D world. During construction, we optimize the solution by detecting 2D correspondences between training views and the corresponding rendered images. We develop a unified differentiable pipeline for camera registration and adjustment of both camera poses and depths, followed by back-projection. We also introduce a novel notion of an expected surface in Gaussian splatting, which is critical to our optimization. These steps enable a coarse solution, which can then be low-pass filtered and refined using standard optimization methods. We demonstrate results on the Tanks and Temples and Static Hikes datasets with as few as three widely-spaced views, showing significantly better quality than competing methods, including those with approximate camera pose information. Moreover, our results improve with more views and outperform previous InstantNGP and Gaussian Splatting algorithms even when using half the dataset. Project page: https://raymondjiangkw.github.io/cogs.github.io/
Authors: Zonglin Lyu, Ming Li, Jianbo Jiao, Chen Chen
Abstract: Recent work in Video Frame Interpolation (VFI) tries to formulate VFI as a diffusion-based conditional image generation problem, synthesizing the intermediate frame given a random noise and neighboring frames. Due to the relatively high resolution of videos, Latent Diffusion Models (LDMs) are employed as the conditional generation model, where the autoencoder compresses images into latent representations for diffusion and then reconstructs images from these latent representations. Such a formulation poses a crucial challenge: VFI expects that the output is deterministically equal to the ground truth intermediate frame, but LDMs randomly generate a diverse set of different images when the model runs multiple times. The reason for the diverse generation is that the cumulative variance (variance accumulated at each step of generation) of generated latent representations in LDMs is large. This makes the sampling trajectory random, resulting in diverse rather than deterministic generations. To address this problem, we propose our unique solution: Frame Interpolation with Consecutive Brownian Bridge Diffusion. Specifically, we propose consecutive Brownian Bridge diffusion that takes a deterministic initial value as input, resulting in a much smaller cumulative variance of generated latent representations. Our experiments suggest that our method can improve together with the improvement of the autoencoder and achieve state-of-the-art performance in VFI, leaving strong potential for further enhancement.
Authors: Jingqun Tang, Qi Liu, Yongjie Ye, Jinghui Lu, Shu Wei, Chunhui Lin, Wanqing Li, Mohamad Fitri Faiz Bin Mahmood, Hao Feng, Zhen Zhao, Yanjie Wang, Yuliang Liu, Hao Liu, Xiang Bai, Can Huang
Abstract: Text-Centric Visual Question Answering (TEC-VQA) in its proper format not only facilitates human-machine interaction in text-centric visual environments but also serves as a de facto gold proxy to evaluate AI models in the domain of text-centric scene understanding. Nonetheless, most existing TEC-VQA benchmarks have focused on high-resource languages like English and Chinese. Despite pioneering works to expand multilingual QA pairs in non-text-centric VQA datasets through translation engines, the translation-based protocol encounters a substantial "visual-textual misalignment" problem when applied to TEC-VQA. Specifically, it prioritizes the text in question-answer pairs while disregarding the visual text present in images. Moreover, it fails to address complexities related to nuanced meaning, contextual distortion, language bias, and question-type diversity. In this work, we tackle multilingual TEC-VQA by introducing MTVQA, the first benchmark featuring high-quality human expert annotations across 9 diverse languages, consisting of 6,778 question-answer pairs across 2,116 images. Further, by comprehensively evaluating numerous state-of-the-art Multimodal Large Language Models (MLLMs), including GPT-4o, GPT-4V, Claude3, and Gemini, on the MTVQA dataset, it is evident that there is still a large room for performance improvement, underscoring the value of MTVQA. Additionally, we supply multilingual training data within the MTVQA dataset, demonstrating that straightforward fine-tuning with this data can substantially enhance multilingual TEC-VQA performance. We aspire that MTVQA will offer the research community fresh insights and stimulate further exploration in multilingual visual text comprehension. The project homepage is available at https://bytedance.github.io/MTVQA/.
Authors: Antoine Legrand, Renaud Detry, Christophe De Vleeschouwer
Abstract: We address the estimation of the 6D pose of an unknown target spacecraft relative to a monocular camera, a key step towards the autonomous rendezvous and proximity operations required by future Active Debris Removal missions. We present a novel method that enables an "off-the-shelf" spacecraft pose estimator, which is supposed to known the target CAD model, to be applied on an unknown target. Our method relies on an in-the wild NeRF, i.e., a Neural Radiance Field that employs learnable appearance embeddings to represent varying illumination conditions found in natural scenes. We train the NeRF model using a sparse collection of images that depict the target, and in turn generate a large dataset that is diverse both in terms of viewpoint and illumination. This dataset is then used to train the pose estimation network. We validate our method on the Hardware-In-the-Loop images of SPEED+ that emulate lighting conditions close to those encountered on orbit. We demonstrate that our method successfully enables the training of an off-the-shelf spacecraft pose estimation network from a sparse set of images. Furthermore, we show that a network trained using our method performs similarly to a model trained on synthetic images generated using the CAD model of the target.
Authors: Hongsheng Wang, Xiang Cai, Xi Sun, Jinhong Yue, Zhanyun Tang, Shengyu Zhang, Feng Lin, Fei Wu
Abstract: Single-view clothed human reconstruction holds a central position in virtual reality applications, especially in contexts involving intricate human motions. It presents notable challenges in achieving realistic clothing deformation. Current methodologies often overlook the influence of motion on surface deformation, resulting in surfaces lacking the constraints imposed by global motion. To overcome these limitations, we introduce an innovative framework, Motion-Based 3D Clo}thed Humans Synthesis (MOSS), which employs kinematic information to achieve motion-aware Gaussian split on the human surface. Our framework consists of two modules: Kinematic Gaussian Locating Splatting (KGAS) and Surface Deformation Detector (UID). KGAS incorporates matrix-Fisher distribution to propagate global motion across the body surface. The density and rotation factors of this distribution explicitly control the Gaussians, thereby enhancing the realism of the reconstructed surface. Additionally, to address local occlusions in single-view, based on KGAS, UID identifies significant surfaces, and geometric reconstruction is performed to compensate for these deformations. Experimental results demonstrate that MOSS achieves state-of-the-art visual quality in 3D clothed human synthesis from monocular videos. Notably, we improve the Human NeRF and the Gaussian Splatting by 33.94% and 16.75% in LPIPS* respectively. Codes are available at https://wanghongsheng01.github.io/MOSS/.
Authors: Boyang Zheng, Jinjin Gu, Shijun Li, Chao Dong
Abstract: The success of large language models (LLMs) has fostered a new research trend of multi-modality large language models (MLLMs), which changes the paradigm of various fields in computer vision. Though MLLMs have shown promising results in numerous high-level vision and vision-language tasks such as VQA and text-to-image, no works have demonstrated how low-level vision tasks can benefit from MLLMs. We find that most current MLLMs are blind to low-level features due to their design of vision modules, thus are inherently incapable for solving low-level vision tasks. In this work, we purpose $\textbf{LM4LV}$, a framework that enables a FROZEN LLM to solve a range of low-level vision tasks without any multi-modal data or prior. This showcases the LLM's strong potential in low-level vision and bridges the gap between MLLMs and low-level vision tasks. We hope this work can inspire new perspectives on LLMs and deeper understanding of their mechanisms. Code is available at https://github.com/bytetriper/LM4LV.
Authors: Yong Liu, Hang Dong, Jinshan Pan, Qingji Dong, Kai Chen, Rongxiang Zhang, Lean Fu, Fei Wang
Abstract: Diffusion models significantly improve the quality of super-resolved images with their impressive content generation capabilities. However, the huge computational costs limit the applications of these methods.Recent efforts have explored reasonable inference acceleration to reduce the number of sampling steps, but the computational cost remains high as each step is performed on the entire image.This paper introduces PatchScaler, a patch-independent diffusion-based single image super-resolution (SR) method, designed to enhance the efficiency of the inference process.The proposed method is motivated by the observation that not all the image patches within an image need the same sampling steps for reconstructing high-resolution images.Based on this observation, we thus develop a Patch-adaptive Group Sampling (PGS) to divide feature patches into different groups according to the patch-level reconstruction difficulty and dynamically assign an appropriate sampling configuration for each group so that the inference speed can be better accelerated.In addition, to improve the denoising ability at each step of the sampling, we develop a texture prompt to guide the estimations of the diffusion model by retrieving high-quality texture priors from a patch-independent reference texture memory.Experiments show that our PatchScaler achieves favorable performance in both quantitative and qualitative evaluations with fast inference speed.Our code and model are available at \url{https://github.com/yongliuy/PatchScaler}.
Authors: Tianyuan Zhang, Lu Wang, Hainan Li, Yisong Xiao, Siyuan Liang, Aishan Liu, Xianglong Liu, Dacheng Tao
Abstract: Lane detection (LD) is an essential component of autonomous driving systems, providing fundamental functionalities like adaptive cruise control and automated lane centering. Existing LD benchmarks primarily focus on evaluating common cases, neglecting the robustness of LD models against environmental illusions such as shadows and tire marks on the road. This research gap poses significant safety challenges since these illusions exist naturally in real-world traffic situations. For the first time, this paper studies the potential threats caused by these environmental illusions to LD and establishes the first comprehensive benchmark LanEvil for evaluating the robustness of LD against this natural corruption. We systematically design 14 prevalent yet critical types of environmental illusions (e.g., shadow, reflection) that cover a wide spectrum of real-world influencing factors in LD tasks. Based on real-world environments, we create 94 realistic and customizable 3D cases using the widely used CARLA simulator, resulting in a dataset comprising 90,292 sampled images. Through extensive experiments, we benchmark the robustness of popular LD methods using LanEvil, revealing substantial performance degradation (-5.37% Accuracy and -10.70% F1-Score on average), with shadow effects posing the greatest risk (-7.39% Accuracy). Additionally, we assess the performance of commercial auto-driving systems OpenPilot and Apollo through collaborative simulations, demonstrating that proposed environmental illusions can lead to incorrect decisions and potential traffic accidents. To defend against environmental illusions, we propose the Attention Area Mixing (AAM) approach using hard examples, which witness significant robustness improvement (+3.76%) under illumination effects. We hope our paper can contribute to advancing more robust auto-driving systems in the future. Website: https://lanevil.github.io/.
Authors: Qilong Zhangli, Jindong Jiang, Di Liu, Licheng Yu, Xiaoliang Dai, Ankit Ramchandani, Guan Pang, Dimitris N. Metaxas, Praveen Krishnan
Abstract: While diffusion models have significantly advanced the quality of image generation, their capability to accurately and coherently render text within these images remains a substantial challenge. Conventional diffusion-based methods for scene text generation are typically limited by their reliance on an intermediate layout output. This dependency often results in a constrained diversity of text styles and fonts, an inherent limitation stemming from the deterministic nature of the layout generation phase. To address these challenges, this paper introduces SceneTextGen, a novel diffusion-based model specifically designed to circumvent the need for a predefined layout stage. By doing so, SceneTextGen facilitates a more natural and varied representation of text. The novelty of SceneTextGen lies in its integration of three key components: a character-level encoder for capturing detailed typographic properties, coupled with a character-level instance segmentation model and a word-level spotting model to address the issues of unwanted text generation and minor character inaccuracies. We validate the performance of our method by demonstrating improved character recognition rates on generated images across different public visual text datasets in comparison to both standard diffusion based methods and text specific methods.
Authors: Junhao Cheng, Xi Lu, Hanhui Li, Khun Loun Zai, Baiqiao Yin, Yuhao Cheng, Yiqiang Yan, Xiaodan Liang
Abstract: As cutting-edge Text-to-Image (T2I) generation models already excel at producing remarkable single images, an even more challenging task, i.e., multi-turn interactive image generation begins to attract the attention of related research communities. This task requires models to interact with users over multiple turns to generate a coherent sequence of images. However, since users may switch subjects frequently, current efforts struggle to maintain subject consistency while generating diverse images. To address this issue, we introduce a training-free multi-agent framework called AutoStudio. AutoStudio employs three agents based on large language models (LLMs) to handle interactions, along with a stable diffusion (SD) based agent for generating high-quality images. Specifically, AutoStudio consists of (i) a subject manager to interpret interaction dialogues and manage the context of each subject, (ii) a layout generator to generate fine-grained bounding boxes to control subject locations, (iii) a supervisor to provide suggestions for layout refinements, and (iv) a drawer to complete image generation. Furthermore, we introduce a Parallel-UNet to replace the original UNet in the drawer, which employs two parallel cross-attention modules for exploiting subject-aware features. We also introduce a subject-initialized generation method to better preserve small subjects. Our AutoStudio hereby can generate a sequence of multi-subject images interactively and consistently. Extensive experiments on the public CMIGBench benchmark and human evaluations show that AutoStudio maintains multi-subject consistency across multiple turns well, and it also raises the state-of-the-art performance by 13.65% in average Frechet Inception Distance and 2.83% in average character-character similarity.
Authors: Tianyi Xiong, Jiayi Wu, Botao He, Cornelia Fermuller, Yiannis Aloimonos, Heng Huang, Christopher A. Metzler
Abstract: By combining differentiable rendering with explicit point-based scene representations, 3D Gaussian Splatting (3DGS) has demonstrated breakthrough 3D reconstruction capabilities. However, to date 3DGS has had limited impact on robotics, where high-speed egomotion is pervasive: Egomotion introduces motion blur and leads to artifacts in existing frame-based 3DGS reconstruction methods. To address this challenge, we introduce Event3DGS, an {\em event-based} 3DGS framework. By exploiting the exceptional temporal resolution of event cameras, Event3GDS can reconstruct high-fidelity 3D structure and appearance under high-speed egomotion. Extensive experiments on multiple synthetic and real-world datasets demonstrate the superiority of Event3DGS compared with existing event-based dense 3D scene reconstruction frameworks; Event3DGS substantially improves reconstruction quality (+3dB) while reducing computational costs by 95\%. Our framework also allows one to incorporate a few motion-blurred frame-based measurements into the reconstruction process to further improve appearance fidelity without loss of structural accuracy.
Authors: Qutub Syed Sha, Michael Paulitsch, Karthik Pattabiraman, Korbinian Hagn, Fabian Oboril, Cornelius Buerkle, Kay-Ulrich Scholl, Gereon Hinz, Alois Knoll
Abstract: As transformer-based object detection models progress, their impact in critical sectors like autonomous vehicles and aviation is expected to grow. Soft errors causing bit flips during inference have significantly impacted DNN performance, altering predictions. Traditional range restriction solutions for CNNs fall short for transformers. This study introduces the Global Clipper and Global Hybrid Clipper, effective mitigation strategies specifically designed for transformer-based models. It significantly enhances their resilience to soft errors and reduces faulty inferences to ~ 0\%. We also detail extensive testing across over 64 scenarios involving two transformer models (DINO-DETR and Lite-DETR) and two CNN models (YOLOv3 and SSD) using three datasets, totalling approximately 3.3 million inferences, to assess model robustness comprehensively. Moreover, the paper explores unique aspects of attention blocks in transformers and their operational differences from CNNs.
Authors: Dominik Scheuble, Chenyang Lei, Seung-Hwan Baek, Mario Bijelic, Felix Heide
Abstract: Lidar has become a cornerstone sensing modality for 3D vision, especially for large outdoor scenarios and autonomous driving. Conventional lidar sensors are capable of providing centimeter-accurate distance information by emitting laser pulses into a scene and measuring the time-of-flight (ToF) of the reflection. However, the polarization of the received light that depends on the surface orientation and material properties is usually not considered. As such, the polarization modality has the potential to improve scene reconstruction beyond distance measurements. In this work, we introduce a novel long-range polarization wavefront lidar sensor (PolLidar) that modulates the polarization of the emitted and received light. Departing from conventional lidar sensors, PolLidar allows access to the raw time-resolved polarimetric wavefronts. We leverage polarimetric wavefronts to estimate normals, distance, and material properties in outdoor scenarios with a novel learned reconstruction method. To train and evaluate the method, we introduce a simulated and real-world long-range dataset with paired raw lidar data, ground truth distance, and normal maps. We find that the proposed method improves normal and distance reconstruction by 53\% mean angular error and 41\% mean absolute error compared to existing shape-from-polarization (SfP) and ToF methods. Code and data are open-sourced at https://light.princeton.edu/pollidar.
Authors: Fangfu Liu, Hanyang Wang, Shunyu Yao, Shengjun Zhang, Jie Zhou, Yueqi Duan
Abstract: In recent years, there has been rapid development in 3D generation models, opening up new possibilities for applications such as simulating the dynamic movements of 3D objects and customizing their behaviors. However, current 3D generative models tend to focus only on surface features such as color and shape, neglecting the inherent physical properties that govern the behavior of objects in the real world. To accurately simulate physics-aligned dynamics, it is essential to predict the physical properties of materials and incorporate them into the behavior prediction process. Nonetheless, predicting the diverse materials of real-world objects is still challenging due to the complex nature of their physical attributes. In this paper, we propose \textbf{Physics3D}, a novel method for learning various physical properties of 3D objects through a video diffusion model. Our approach involves designing a highly generalizable physical simulation system based on a viscoelastic material model, which enables us to simulate a wide range of materials with high-fidelity capabilities. Moreover, we distill the physical priors from a video diffusion model that contains more understanding of realistic object materials. Extensive experiments demonstrate the effectiveness of our method with both elastic and plastic materials. Physics3D shows great potential for bridging the gap between the physical world and virtual neural space, providing a better integration and application of realistic physical principles in virtual environments. Project page: https://liuff19.github.io/Physics3D.
Authors: Yue Lu, Shizhou Zhang, De Cheng, Yinghui Xing, Nannan Wang, Peng Wang, Yanning Zhang
Abstract: Existing prompt-tuning methods have demonstrated impressive performances in continual learning (CL), by selecting and updating relevant prompts in the vision-transformer models. On the contrary, this paper aims to learn each task by tuning the prompts in the direction orthogonal to the subspace spanned by previous tasks' features, so as to ensure no interference on tasks that have been learned to overcome catastrophic forgetting in CL. However, different from the orthogonal projection in the traditional CNN architecture, the prompt gradient orthogonal projection in the ViT architecture shows completely different and greater challenges, i.e., 1) the high-order and non-linear self-attention operation; 2) the drift of prompt distribution brought by the LayerNorm in the transformer block. Theoretically, we have finally deduced two consistency conditions to achieve the prompt gradient orthogonal projection, which provide a theoretical guarantee of eliminating interference on previously learned knowledge via the self-attention mechanism in visual prompt tuning. In practice, an effective null-space-based approximation solution has been proposed to implement the prompt gradient orthogonal projection. Extensive experimental results demonstrate the effectiveness of anti-forgetting on four class-incremental benchmarks with diverse pre-trained baseline models, and our approach achieves superior performances to state-of-the-art methods. Our code is available at https://github.com/zugexiaodui/VPTinNSforCL.
Authors: Hou-I Liu, Yu-Wen Tseng, Kai-Cheng Chang, Pin-Jyun Wang, Hong-Han Shuai, Wen-Huang Cheng
Abstract: Despite notable advancements in the field of computer vision, the precise detection of tiny objects continues to pose a significant challenge, largely owing to the minuscule pixel representation allocated to these objects in imagery data. This challenge resonates profoundly in the domain of geoscience and remote sensing, where high-fidelity detection of tiny objects can facilitate a myriad of applications ranging from urban planning to environmental monitoring. In this paper, we propose a new framework, namely, DeNoising FPN with Trans R-CNN (DNTR), to improve the performance of tiny object detection. DNTR consists of an easy plug-in design, DeNoising FPN (DN-FPN), and an effective Transformer-based detector, Trans R-CNN. Specifically, feature fusion in the feature pyramid network is important for detecting multiscale objects. However, noisy features may be produced during the fusion process since there is no regularization between the features of different scales. Therefore, we introduce a DN-FPN module that utilizes contrastive learning to suppress noise in each level's features in the top-down path of FPN. Second, based on the two-stage framework, we replace the obsolete R-CNN detector with a novel Trans R-CNN detector to focus on the representation of tiny objects with self-attention. Experimental results manifest that our DNTR outperforms the baselines by at least 17.4% in terms of APvt on the AI-TOD dataset and 9.6% in terms of AP on the VisDrone dataset, respectively. Our code will be available at https://github.com/hoiliu-0801/DNTR.
Authors: Qingsong Xie, Zhenyi Liao, Zhijie Deng, Chen chen, Shixiang Tang, Haonan Lu
Abstract: Distilling large latent diffusion models (LDMs) into ones that are fast to sample from is attracting growing research interest. However, the majority of existing methods face a dilemma where they either (i) depend on multiple individual distilled models for different sampling budgets, or (ii) sacrifice generation quality with limited (e.g., 2-4) and/or moderate (e.g., 5-8) sampling steps. To address these, we extend the recent multistep consistency distillation (MCD) strategy to representative LDMs, establishing the Multistep Latent Consistency Models (MLCMs) approach for low-cost high-quality image synthesis. MLCM serves as a unified model for various sampling steps due to the promise of MCD. We further augment MCD with a progressive training strategy to strengthen inter-segment consistency to boost the quality of few-step generations. We take the states from the sampling trajectories of the teacher model as training data for MLCMs to lift the requirements for high-quality training datasets and to bridge the gap between the training and inference of the distilled model. MLCM is compatible with preference learning strategies for further improvement of visual quality and aesthetic appeal. Empirically, MLCM can generate high-quality, delightful images with only 2-8 sampling steps. On the MSCOCO-2017 5K benchmark, MLCM distilled from SDXL gets a CLIP Score of 33.30, Aesthetic Score of 6.19, and Image Reward of 1.20 with only 4 steps, substantially surpassing 4-step LCM [23], 8-step SDXL-Lightning [17], and 8-step HyperSD [33]. We also demonstrate the versatility of MLCMs in applications including controllable generation, image style transfer, and Chinese-to-image generation.
Authors: Abdulwahab Felemban, Eslam Mohamed Bakr, Xiaoqian Shen, Jian Ding, Abduallah Mohamed, Mohamed Elhoseiny
Abstract: We introduce iMotion-LLM: a Multimodal Large Language Models (LLMs) with trajectory prediction, tailored to guide interactive multi-agent scenarios. Different from conventional motion prediction approaches, iMotion-LLM capitalizes on textual instructions as key inputs for generating contextually relevant trajectories. By enriching the real-world driving scenarios in the Waymo Open Dataset with textual motion instructions, we created InstructWaymo. Leveraging this dataset, iMotion-LLM integrates a pretrained LLM, fine-tuned with LoRA, to translate scene features into the LLM input space. iMotion-LLM offers significant advantages over conventional motion prediction models. First, it can generate trajectories that align with the provided instructions if it is a feasible direction. Second, when given an infeasible direction, it can reject the instruction, thereby enhancing safety. These findings act as milestones in empowering autonomous navigation systems to interpret and predict the dynamics of multi-agent environments, laying the groundwork for future advancements in this field.
Authors: Xiaojie Li, Chenghao Gu, Shuzhao Xie, Yunpeng Bai, Weixiang Zhang, Zhi Wang
Abstract: Fine-Tuning Diffusion Models enable a wide range of personalized generation and editing applications on diverse visual modalities. While Low-Rank Adaptation (LoRA) accelerates the fine-tuning process, it still requires multiple reference images and time-consuming training, which constrains its scalability for large-scale and real-time applications. In this paper, we propose \textit{View Iterative Self-Attention Control (VisCtrl)} to tackle this challenge. Specifically, VisCtrl is a training-free method that injects the appearance and structure of a user-specified subject into another subject in the target image, unlike previous approaches that require fine-tuning the model. Initially, we obtain the initial noise for both the reference and target images through DDIM inversion. Then, during the denoising phase, features from the reference image are injected into the target image via the self-attention mechanism. Notably, by iteratively performing this feature injection process, we ensure that the reference image features are gradually integrated into the target image. This approach results in consistent and harmonious editing with only one reference image in a few denoising steps. Moreover, benefiting from our plug-and-play architecture design and the proposed Feature Gradual Sampling strategy for multi-view editing, our method can be easily extended to edit in complex visual domains. Extensive experiments show the efficacy of VisCtrl across a spectrum of tasks, including personalized editing of images, videos, and 3D scenes.
Authors: Simon Dahan, Logan Z. J. Williams, Daniel Rueckert, Emma C. Robinson
Abstract: Surface meshes are a favoured domain for representing structural and functional information on the human cortex, but their complex topology and geometry pose significant challenges for deep learning analysis. While Transformers have excelled as domain-agnostic architectures for sequence-to-sequence learning, the quadratic cost of the self-attention operation remains an obstacle for many dense prediction tasks. Inspired by some of the latest advances in hierarchical modelling with vision transformers, we introduce the Multiscale Surface Vision Transformer (MS-SiT) as a backbone architecture for surface deep learning. The self-attention mechanism is applied within local-mesh-windows to allow for high-resolution sampling of the underlying data, while a shifted-window strategy improves the sharing of information between windows. Neighbouring patches are successively merged, allowing the MS-SiT to learn hierarchical representations suitable for any prediction task. Results demonstrate that the MS-SiT outperforms existing surface deep learning methods for neonatal phenotyping prediction tasks using the Developing Human Connectome Project (dHCP) dataset. Furthermore, building the MS-SiT backbone into a U-shaped architecture for surface segmentation demonstrates competitive results on cortical parcellation using the UK Biobank (UKB) and manually-annotated MindBoggle datasets. Code and trained models are publicly available at https://github.com/metrics-lab/surface-vision-transformers.
URLs: https://github.com/metrics-lab/surface-vision-transformers.
Authors: Young D. Kwon, Rui Li, Stylianos I. Venieris, Jagmohan Chauhan, Nicholas D. Lane, Cecilia Mascolo
Abstract: On-device training is essential for user personalisation and privacy. With the pervasiveness of IoT devices and microcontroller units (MCUs), this task becomes more challenging due to the constrained memory and compute resources, and the limited availability of labelled user data. Nonetheless, prior works neglect the data scarcity issue, require excessively long training time (e.g. a few hours), or induce substantial accuracy loss (>10%). In this paper, we propose TinyTrain, an on-device training approach that drastically reduces training time by selectively updating parts of the model and explicitly coping with data scarcity. TinyTrain introduces a task-adaptive sparse-update method that dynamically selects the layer/channel to update based on a multi-objective criterion that jointly captures user data, the memory, and the compute capabilities of the target device, leading to high accuracy on unseen tasks with reduced computation and memory footprint. TinyTrain outperforms vanilla fine-tuning of the entire network by 3.6-5.0% in accuracy, while reducing the backward-pass memory and computation cost by up to 1,098x and 7.68x, respectively. Targeting broadly used real-world edge devices, TinyTrain achieves 9.5x faster and 3.5x more energy-efficient training over status-quo approaches, and 2.23x smaller memory footprint than SOTA methods, while remaining within the 1 MB memory envelope of MCU-grade platforms.
Authors: Simon Dahan, Logan Z. J. Williams, Yourong Guo, Daniel Rueckert, Emma C. Robinson
Abstract: The development of robust and generalisable models for encoding the spatio-temporal dynamics of human brain activity is crucial for advancing neuroscientific discoveries. However, significant individual variation in the organisation of the human cerebral cortex makes it difficult to identify population-level trends in these signals. Recently, Surface Vision Transformers (SiTs) have emerged as a promising approach for modelling cortical signals, yet they face some limitations in low-data scenarios due to the lack of inductive biases in their architecture. To address these challenges, this paper proposes the surface Masked AutoEncoder (sMAE) and video surface Masked AutoEncoder (vsMAE) - for multivariate and spatio-temporal pre-training of cortical signals over regular icosahedral grids. These models are trained to reconstruct cortical feature maps from masked versions of the input by learning strong latent representations of cortical structure and function. Such representations translate into better modelling of individual phenotypes and enhanced performance in downstream tasks. The proposed approach was evaluated on cortical phenotype regression using data from the young adult Human Connectome Project (HCP) and developing HCP (dHCP). Results show that (v)sMAE pre-trained models improve phenotyping prediction performance on multiple tasks by $\ge 26\%$, and offer faster convergence relative to models trained from scratch. Finally, we show that pre-training vision transformers on large datasets, such as the UK Biobank (UKB), supports transfer learning to low-data regimes. Our code and pre-trained models are publicly available at https://github.com/metrics-lab/surface-masked-autoencoders .
URLs: https://github.com/metrics-lab/surface-masked-autoencoders
Authors: Wenhan Yang, Jingdong Gao, Baharan Mirzasoleiman
Abstract: Contrastive Language-Image Pre-training (CLIP) on large image-caption datasets has achieved remarkable success in zero-shot classification and enabled transferability to new domains. However, CLIP is extremely more vulnerable to targeted data poisoning and backdoor attacks, compared to supervised learning. Perhaps surprisingly, poisoning 0.0001% of CLIP pre-training data is enough to make targeted data poisoning attacks successful. This is four orders of magnitude smaller than what is required to poison supervised models. Despite this vulnerability, existing methods are very limited in defending CLIP models during pre-training. In this work, we propose a strong defense, SAFECLIP, to safely pre-train CLIP against targeted data poisoning and backdoor attacks. SAFECLIP warms up the model by applying unimodal contrastive learning (CL) on image and text modalities separately. Then, it divides the data into safe and risky sets, by applying a Gaussian Mixture Model to the cosine similarity of image-caption pair representations. SAFECLIP pre-trains the model by applying the CLIP loss to the safe set and applying unimodal CL to image and text modalities of the risky set separately. By gradually increasing the size of the safe set during pre-training, SAFECLIP effectively breaks targeted data poisoning and backdoor attacks without harming the CLIP performance. Our extensive experiments on CC3M, Visual Genome, and MSCOCO demonstrate that SAFECLIP significantly reduces the success rate of targeted data poisoning attacks from 93.75% to 0% and that of various backdoor attacks from up to 100% to 0%, without harming CLIP's performance.
Authors: Hao Dong, Ga\"etan Frusque, Yue Zhao, Eleni Chatzi, Olga Fink
Abstract: Anomaly detection (AD) is essential in identifying rare and often critical events in complex systems, finding applications in fields such as network intrusion detection, financial fraud detection, and fault detection in infrastructure and industrial systems. While AD is typically treated as an unsupervised learning task due to the high cost of label annotation, it is more practical to assume access to a small set of labeled anomaly samples from domain experts, as is the case for semi-supervised anomaly detection. Semi-supervised and supervised approaches can leverage such labeled data, resulting in improved performance. In this paper, rather than proposing a new semi-supervised or supervised approach for AD, we introduce a novel algorithm for generating additional pseudo-anomalies on the basis of the limited labeled anomalies and a large volume of unlabeled data. This serves as an augmentation to facilitate the detection of new anomalies. Our proposed algorithm, named Nearest Neighbor Gaussian Mixup (NNG-Mix), efficiently integrates information from both labeled and unlabeled data to generate pseudo-anomalies. We compare the performance of this novel algorithm with commonly applied augmentation techniques, such as Mixup and Cutout. We evaluate NNG-Mix by training various existing semi-supervised and supervised anomaly detection algorithms on the original training data along with the generated pseudo-anomalies. Through extensive experiments on 57 benchmark datasets in ADBench, reflecting different data types, we demonstrate that NNG-Mix outperforms other data augmentation methods. It yields significant performance improvements compared to the baselines trained exclusively on the original training data. Notably, NNG-Mix yields up to 16.4%, 8.8%, and 8.0% improvements on Classical, CV, and NLP datasets in ADBench. Our source code is available at https://github.com/donghao51/NNG-Mix.
Authors: Shudong Li, Xiao Jiang, Matthew Tivnan, Grace J. Gang, Yuan Shen, J. Webster Stayman
Abstract: Diffusion models have been demonstrated as powerful deep learning tools for image generation in CT reconstruction and restoration. Recently, diffusion posterior sampling, where a score-based diffusion prior is combined with a likelihood model, has been used to produce high quality CT images given low-quality measurements. This technique is attractive since it permits a one-time, unsupervised training of a CT prior; which can then be incorporated with an arbitrary data model. However, current methods rely on a linear model of x-ray CT physics to reconstruct or restore images. While it is common to linearize the transmission tomography reconstruction problem, this is an approximation to the true and inherently nonlinear forward model. We propose a new method that solves the inverse problem of nonlinear CT image reconstruction via diffusion posterior sampling. We implement a traditional unconditional diffusion model by training a prior score function estimator, and apply Bayes rule to combine this prior with a measurement likelihood score function derived from the nonlinear physical model to arrive at a posterior score function that can be used to sample the reverse-time diffusion process. This plug-and-play method allows incorporation of a diffusion-based prior with generalized nonlinear CT image reconstruction into multiple CT system designs with different forward models, without the need for any additional training. We develop the algorithm that performs this reconstruction, including an ordered-subsets variant for accelerated processing and demonstrate the technique in both fully sampled low dose data and sparse-view geometries using a single unsupervised training of the prior.
Authors: Jonas Dippel, Barbara Feulner, Tobias Winterhoff, Timo Milbich, Stephan Tietz, Simon Schallenberg, Gabriel Dernbach, Andreas Kunft, Simon Heinke, Marie-Lisa Eich, Julika Ribbat-Idel, Rosemarie Krupar, Philipp Anders, Niklas Preni{\ss}l, Philipp Jurmeister, David Horst, Lukas Ruff, Klaus-Robert M\"uller, Frederick Klauschen, Maximilian Alber
Abstract: Artificial intelligence has started to transform histopathology impacting clinical diagnostics and biomedical research. However, while many computational pathology approaches have been proposed, most current AI models are limited with respect to generalization, application variety, and handling rare diseases. Recent efforts introduced self-supervised foundation models to address these challenges, yet existing approaches do not leverage pathologist knowledge by design. In this study, we present a novel approach to designing foundation models for computational pathology, incorporating pathologist expertise, semi-automated data curation, and a diverse dataset from over 15 laboratories, including 58 tissue types, and encompassing 129 different histochemical and immunohistochemical staining modalities. We demonstrate that our model "RudolfV" surpasses existing state-of-the-art foundation models across different benchmarks focused on tumor microenvironment profiling, biomarker evaluation, and reference case search while exhibiting favorable robustness properties. Our study shows how domain-specific knowledge can increase the efficiency and performance of pathology foundation models and enable novel application areas.
Authors: Siwei Wu, Yizhi Li, Kang Zhu, Ge Zhang, Yiming Liang, Kaijing Ma, Chenghao Xiao, Haoran Zhang, Bohao Yang, Wenhu Chen, Wenhao Huang, Noura Al Moubayed, Jie Fu, Chenghua Lin
Abstract: Multi-modal information retrieval (MMIR) is a rapidly evolving field, where significant progress, particularly in image-text pairing, has been made through advanced representation learning and cross-modality alignment research. However, current benchmarks for evaluating MMIR performance in image-text pairing within the scientific domain show a notable gap, where chart and table images described in scholarly language usually do not play a significant role. To bridge this gap, we develop a specialised scientific MMIR (SciMMIR) benchmark by leveraging open-access paper collections to extract data relevant to the scientific domain. This benchmark comprises 530K meticulously curated image-text pairs, extracted from figures and tables with detailed captions in scientific documents. We further annotate the image-text pairs with two-level subset-subcategory hierarchy annotations to facilitate a more comprehensive evaluation of the baselines. We conducted zero-shot and fine-tuning evaluations on prominent multi-modal image-captioning and visual language models, such as CLIP and BLIP. Our analysis offers critical insights for MMIR in the scientific domain, including the impact of pre-training and fine-tuning settings and the influence of the visual and textual encoders. All our data and checkpoints are publicly available at https://github.com/Wusiwei0410/SciMMIR.
Authors: Dongping Chen, Ruoxi Chen, Shilin Zhang, Yinuo Liu, Yaochen Wang, Huichi Zhou, Qihui Zhang, Yao Wan, Pan Zhou, Lichao Sun
Abstract: Multimodal Large Language Models (MLLMs) have gained significant attention recently, showing remarkable potential in artificial general intelligence. However, assessing the utility of MLLMs presents considerable challenges, primarily due to the absence of multimodal benchmarks that align with human preferences. Drawing inspiration from the concept of LLM-as-a-Judge within LLMs, this paper introduces a novel benchmark, termed MLLM-as-a-Judge, to assess the ability of MLLMs in assisting judges across diverse modalities, encompassing three distinct tasks: Scoring Evaluation, Pair Comparison, and Batch Ranking. Our study reveals that, while MLLMs demonstrate remarkable human-like discernment in Pair Comparison, there is a significant divergence from human preferences in Scoring Evaluation and Batch Ranking. Furthermore, a closer examination reveals persistent challenges in the judgment capacities of LLMs, including diverse biases, hallucinatory responses, and inconsistencies in judgment, even in advanced models such as GPT-4V. These findings emphasize the pressing need for enhancements and further research efforts to be undertaken before regarding MLLMs as fully reliable evaluators. In light of this, we advocate for additional efforts dedicated to supporting the continuous development within the domain of MLLM functioning as judges. The code and dataset are publicly available at our project homepage: \url{https://mllm-judge.github.io/}.
Authors: Moran Yanuka, Morris Alper, Hadar Averbuch-Elor, Raja Giryes
Abstract: Web-scale training on paired text-image data is becoming increasingly central to multimodal learning, but is challenged by the highly noisy nature of datasets in the wild. Standard data filtering approaches succeed in removing mismatched text-image pairs, but permit semantically related but highly abstract or subjective text. These approaches lack the fine-grained ability to isolate the most concrete samples that provide the strongest signal for learning in a noisy dataset. In this work, we propose a new metric, image caption concreteness, that evaluates caption text without an image reference to measure its concreteness and relevancy for use in multimodal learning. Our approach leverages strong foundation models for measuring visual-semantic information loss in multimodal representations. We demonstrate that this strongly correlates with human evaluation of concreteness in both single-word and sentence-level texts. Moreover, we show that curation using ICC complements existing approaches: It succeeds in selecting the highest quality samples from multimodal web-scale datasets to allow for efficient training in resource-constrained settings.
Authors: Kirill Muravyev, Alexander Melekhin, Dmitry Yudin, Konstantin Yakovlev
Abstract: Mapping is one of the crucial tasks enabling autonomous navigation of a mobile robot. Conventional mapping methods output dense geometric map representation, e.g. an occupancy grid, which is not trivial to keep consistent for the prolonged runs covering large environments. Meanwhile, capturing the topological structure of the workspace enables fast path planning, is less prone to odometry error accumulation and does not consume much memory. Following this idea, this paper introduces PRISM-TopoMap -- a topological mapping method that maintains a graph of locally aligned locations not relying on global metric coordinates. The proposed method involves learnable multimodal place recognition paired with the scan matching pipeline for localization and loop closure in the graph of locations. The latter is updated online and the robot is localized in a proper node at each time step. We conduct a broad experimental evaluation of the suggested approach in a range of photo-realistic environments and on a real robot (wheeled differential driven Husky robot), and compare it to state of the art. The results of the empirical evaluation confirm that PRISM-Topomap consistently outperforms competitors across several measures of mapping and navigation efficiency and performs well on a real robot. The code of PRISM-Topomap is open-sourced and available at https://github.com/kirillMouraviev/prism-topomap.
Authors: Jordan Dotzel, Yuzong Chen, Bahaa Kotb, Sushma Prasad, Gang Wu, Sheng Li, Mohamed S. Abdelfattah, Zhiru Zhang
Abstract: The increasing size of large language models (LLMs) traditionally requires low-precision integer formats to meet strict latency and power demands. Yet recently, alternative formats such as Normal Float (NF4) have increased model accuracy at the cost of increased chip area. In this work, we first conduct a large-scale analysis of LLM weights and activations across 30 networks and conclude that most distributions follow a Student's t-distribution. We then derive a new theoretically optimal format, Student Float (SF4), that improves over NF4 across modern LLMs, for example increasing the average accuracy on LLaMA2-7B by 0.76% across tasks. Using this format as a high-accuracy reference, we then propose augmenting E2M1 with two variants of supernormal support for higher model accuracy. Finally, we explore the quality and efficiency frontier across 11 datatypes by evaluating their model accuracy and hardware complexity. We discover a Pareto curve composed of INT4, E2M1, and E2M1 with supernormal support, which offers a continuous tradeoff between model accuracy and chip area. For example, E2M1 with supernormal support increases the accuracy of Phi-2 by up to 2.19% with 1.22% area overhead, enabling more LLM-based applications to be run at four bits. The supporting code is hosted at https://github.com/cornell-zhang/llm-datatypes.
Authors: Jiashu Xu
Abstract: Automatic medical image segmentation technology has the potential to expedite pathological diagnoses, thereby enhancing the efficiency of patient care. However, medical images often have complex textures and structures, and the models often face the problem of reduced image resolution and information loss due to downsampling. To address this issue, we propose HC-Mamba, a new medical image segmentation model based on the modern state space model Mamba. Specifically, we introduce the technique of dilated convolution in the HC-Mamba model to capture a more extensive range of contextual information without increasing the computational cost by extending the perceptual field of the convolution kernel. In addition, the HC-Mamba model employs depthwise separable convolutions, significantly reducing the number of parameters and the computational power of the model. By combining dilated convolution and depthwise separable convolutions, HC-Mamba is able to process large-scale medical image data at a much lower computational cost while maintaining a high level of performance. We conduct comprehensive experiments on segmentation tasks including organ segmentation and skin lesion, and conduct extensive experiments on Synapse, ISIC17 and ISIC18 to demonstrate the potential of the HC-Mamba model in medical image segmentation. The experimental results show that HC-Mamba exhibits competitive performance on all these datasets, thereby proving its effectiveness and usefulness in medical image segmentation.
Authors: Xiaosong Jia, Zhenjie Yang, Qifeng Li, Zhiyuan Zhang, Junchi Yan
Abstract: In an era marked by the rapid scaling of foundation models, autonomous driving technologies are approaching a transformative threshold where end-to-end autonomous driving (E2E-AD) emerges due to its potential of scaling up in the data-driven manner. However, existing E2E-AD methods are mostly evaluated under the open-loop log-replay manner with L2 errors and collision rate as metrics (e.g., in nuScenes), which could not fully reflect the driving performance of algorithms as recently acknowledged in the community. For those E2E-AD methods evaluated under the closed-loop protocol, they are tested in fixed routes (e.g., Town05Long and Longest6 in CARLA) with the driving score as metrics, which is known for high variance due to the unsmoothed metric function and large randomness in the long route. Besides, these methods usually collect their own data for training, which makes algorithm-level fair comparison infeasible. To fulfill the paramount need of comprehensive, realistic, and fair testing environments for Full Self-Driving (FSD), we present Bench2Drive, the first benchmark for evaluating E2E-AD systems' multiple abilities in a closed-loop manner. Bench2Drive's official training data consists of 2 million fully annotated frames, collected from 10000 short clips uniformly distributed under 44 interactive scenarios (cut-in, overtaking, detour, etc), 23 weathers (sunny, foggy, rainy, etc), and 12 towns (urban, village, university, etc) in CARLA v2. Its evaluation protocol requires E2E-AD models to pass 44 interactive scenarios under different locations and weathers which sums up to 220 routes and thus provides a comprehensive and disentangled assessment about their driving capability under different situations. We implement state-of-the-art E2E-AD models and evaluate them in Bench2Drive, providing insights regarding current status and future directions.
Authors: David R Wessels, David M Knigge, Samuele Papa, Riccardo Valperga, Sharvaree Vadgama, Efstratios Gavves, Erik J Bekkers
Abstract: Recently, Neural Fields have emerged as a powerful modelling paradigm to represent continuous signals. In a conditional neural field, a field is represented by a latent variable that conditions the NeF, whose parametrisation is otherwise shared over an entire dataset. We propose Equivariant Neural Fields based on cross attention transformers, in which NeFs are conditioned on a geometric conditioning variable, a latent point cloud, that enables an equivariant decoding from latent to field. Our equivariant approach induces a steerability property by which both field and latent are grounded in geometry and amenable to transformation laws if the field transforms, the latent represents transforms accordingly and vice versa. Crucially, the equivariance relation ensures that the latent is capable of (1) representing geometric patterns faitfhully, allowing for geometric reasoning in latent space, (2) weightsharing over spatially similar patterns, allowing for efficient learning of datasets of fields. These main properties are validated using classification experiments and a verification of the capability of fitting entire datasets, in comparison to other non-equivariant NeF approaches. We further validate the potential of ENFs by demonstrate unique local field editing properties.
Authors: Ziqiang Liu, Feiteng Fang, Xi Feng, Xinrun Du, Chenhao Zhang, Zekun Wang, Yuelin Bai, Qixuan Zhao, Liyang Fan, Chengguang Gan, Hongquan Lin, Jiaming Li, Yuansheng Ni, Haihong Wu, Yaswanth Narsupalli, Zhigang Zheng, Chengming Li, Xiping Hu, Ruifeng Xu, Xiaojun Chen, Min Yang, Jiaheng Liu, Ruibo Liu, Wenhao Huang, Ge Zhang, Shiwen Ni
Abstract: The rapid advancements in the development of multimodal large language models (MLLMs) have consistently led to new breakthroughs on various benchmarks. In response, numerous challenging and comprehensive benchmarks have been proposed to more accurately assess the capabilities of MLLMs. However, there is a dearth of exploration of the higher-order perceptual capabilities of MLLMs. To fill this gap, we propose the Image Implication understanding Benchmark, II-Bench, which aims to evaluate the model's higher-order perception of images. Through extensive experiments on II-Bench across multiple MLLMs, we have made significant findings. Initially, a substantial gap is observed between the performance of MLLMs and humans on II-Bench. The pinnacle accuracy of MLLMs attains 74.8%, whereas human accuracy averages 90%, peaking at an impressive 98%. Subsequently, MLLMs perform worse on abstract and complex images, suggesting limitations in their ability to understand high-level semantics and capture image details. Finally, it is observed that most models exhibit enhanced accuracy when image sentiment polarity hints are incorporated into the prompts. This observation underscores a notable deficiency in their inherent understanding of image sentiment. We believe that II-Bench will inspire the community to develop the next generation of MLLMs, advancing the journey towards expert artificial general intelligence (AGI). II-Bench is publicly available at https://huggingface.co/datasets/m-a-p/II-Bench.