Authors: Yoav HaCohen, Nisan Chiprut, Benny Brazowski, Daniel Shalem, Dudu Moshe, Eitan Richardson, Eran Levin, Guy Shiran, Nir Zabari, Ori Gordon, Poriya Panet, Sapir Weissbuch, Victor Kulikov, Yaki Bitterman, Zeev Melumian, Ofir Bibi
Abstract: We introduce LTX-Video, a transformer-based latent diffusion model that adopts a holistic approach to video generation by seamlessly integrating the responsibilities of the Video-VAE and the denoising transformer. Unlike existing methods, which treat these components as independent, LTX-Video aims to optimize their interaction for improved efficiency and quality. At its core is a carefully designed Video-VAE that achieves a high compression ratio of 1:192, with spatiotemporal downscaling of 32 x 32 x 8 pixels per token, enabled by relocating the patchifying operation from the transformer's input to the VAE's input. Operating in this highly compressed latent space enables the transformer to efficiently perform full spatiotemporal self-attention, which is essential for generating high-resolution videos with temporal consistency. However, the high compression inherently limits the representation of fine details. To address this, our VAE decoder is tasked with both latent-to-pixel conversion and the final denoising step, producing the clean result directly in pixel space. This approach preserves the ability to generate fine details without incurring the runtime cost of a separate upsampling module. Our model supports diverse use cases, including text-to-video and image-to-video generation, with both capabilities trained simultaneously. It achieves faster-than-real-time generation, producing 5 seconds of 24 fps video at 768x512 resolution in just 2 seconds on an Nvidia H100 GPU, outperforming all existing models of similar scale. The source code and pre-trained models are publicly available, setting a new benchmark for accessible and scalable video generation.
Authors: Xiaozhou You, Jian Zhang
Abstract: Generating desired images conditioned on given text descriptions has received lots of attention. Recently, diffusion models and autoregressive models have demonstrated their outstanding expressivity and gradually replaced GAN as the favored architectures for text-to-image synthesis. However, they still face some obstacles: slow inference speed and expensive training costs. To achieve more powerful and faster text-to-image synthesis under complex scenes, we propose TIGER, a text-to-image GAN with pretrained representations. To be specific, we propose a vision-empowered discriminator and a high-capacity generator. (i) The vision-empowered discriminator absorbs the complex scene understanding ability and the domain generalization ability from pretrained vision models to enhance model performance. Unlike previous works, we explore stacking multiple pretrained models in our discriminator to collect multiple different representations. (ii) The high-capacity generator aims to achieve effective text-image fusion while increasing the model capacity. The high-capacity generator consists of multiple novel high-capacity fusion blocks (HFBlock). And the HFBlock contains several deep fusion modules and a global fusion module, which play different roles to benefit our model. Extensive experiments demonstrate the outstanding performance of our proposed TIGER both on standard and zero-shot text-to-image synthesis tasks. On the standard text-to-image synthesis task, TIGER achieves state-of-the-art performance on two challenging datasets, which obtain a new FID 5.48 (COCO) and 9.38 (CUB). On the zero-shot text-to-image synthesis task, we achieve comparable performance with fewer model parameters, smaller training data size and faster inference speed. Additionally, more experiments and analyses are conducted in the Supplementary Material.
Authors: Jiaojiao Ye, Zhen Wang, Linnan Jiang
Abstract: Diffusionmodels(DMs)havedemonstratedremarkableachievements in synthesizing images of high fidelity and diversity. However, the extensive computational requirements and slow generative speed of diffusion models have limited their widespread adoption. In this paper, we propose a novel post-training quantization for diffusion models (PQD), which is a time-aware optimization framework for diffusion models based on post-training quantization. The proposed framework optimizes the inference process by selecting representative samples and conducting time-aware calibration. Experimental results show that our proposed method is able to directly quantize full-precision diffusion models into 8-bit or 4-bit models while maintaining comparable performance in a training-free manner, achieving a few FID change on ImageNet for unconditional image generation. Our approach demonstrates compatibility and can also be applied to 512x512 text-guided image generation for the first time.
Authors: Taniya Das, Louis Mahon, Thomas Lukasiewicz
Abstract: One of the challenging tasks in the field of video understanding is extracting semantic content from video inputs. Most existing systems use language models to describe videos in natural language sentences, but this has several major shortcomings. Such systems can rely too heavily on the language model component and base their output on statistical regularities in natural language text rather than on the visual contents of the video. Additionally, natural language annotations cannot be readily processed by a computer, are difficult to evaluate with performance metrics and cannot be easily translated into a different natural language. In this paper, we propose a method to annotate videos with knowledge graphs, and so avoid these problems. Specifically, we propose a deep-learning-based model for this task that first predicts pairs of individuals and then the relations between them. Additionally, we propose an extension of our model for the inclusion of background knowledge in the construction of knowledge graphs.
Authors: Jeremy Klotz, Shree K. Nayar
Abstract: A minimalist vision system uses the smallest number of pixels needed to solve a vision task. While traditional cameras use a large grid of square pixels, a minimalist camera uses freeform pixels that can take on arbitrary shapes to increase their information content. We show that the hardware of a minimalist camera can be modeled as the first layer of a neural network, where the subsequent layers are used for inference. Training the network for any given task yields the shapes of the camera's freeform pixels, each of which is implemented using a photodetector and an optical mask. We have designed minimalist cameras for monitoring indoor spaces (with 8 pixels), measuring room lighting (with 8 pixels), and estimating traffic flow (with 8 pixels). The performance demonstrated by these systems is on par with a traditional camera with orders of magnitude more pixels. Minimalist vision has two major advantages. First, it naturally tends to preserve the privacy of individuals in the scene since the captured information is inadequate for extracting visual details. Second, since the number of measurements made by a minimalist camera is very small, we show that it can be fully self-powered, i.e., function without an external power supply or a battery.
Authors: Zhenting Wang, Shuming Hu, Shiyu Zhao, Xiaowen Lin, Felix Juefei-Xu, Zhuowei Li, Ligong Han, Harihar Subramanyam, Li Chen, Jianfa Chen, Nan Jiang, Lingjuan Lyu, Shiqing Ma, Dimitris N. Metaxas, Ankit Jain
Abstract: Image content safety has become a significant challenge with the rise of visual media on online platforms. Meanwhile, in the age of AI-generated content (AIGC), many image generation models are capable of producing harmful content, such as images containing sexual or violent material. Thus, it becomes crucial to identify such unsafe images based on established safety rules. Pre-trained Multimodal Large Language Models (MLLMs) offer potential in this regard, given their strong pattern recognition abilities. Existing approaches typically fine-tune MLLMs with human-labeled datasets, which however brings a series of drawbacks. First, relying on human annotators to label data following intricate and detailed guidelines is both expensive and labor-intensive. Furthermore, users of safety judgment systems may need to frequently update safety rules, making fine-tuning on human-based annotation more challenging. This raises the research question: Can we detect unsafe images by querying MLLMs in a zero-shot setting using a predefined safety constitution (a set of safety rules)? Our research showed that simply querying pre-trained MLLMs does not yield satisfactory results. This lack of effectiveness stems from factors such as the subjectivity of safety rules, the complexity of lengthy constitutions, and the inherent biases in the models. To address these challenges, we propose a MLLM-based method includes objectifying safety rules, assessing the relevance between rules and images, making quick judgments based on debiased token probabilities with logically complete yet simplified precondition chains for safety rules, and conducting more in-depth reasoning with cascaded chain-of-thought processes if necessary. Experiment results demonstrate that our method is highly effective for zero-shot image safety judgment tasks.
Authors: Zixuan Yin, Han Sun, Ningzhong Liu, Huiyu Zhou, Jiaquan Shen
Abstract: Lidars and cameras play essential roles in autonomous driving, offering complementary information for 3D detection. The state-of-the-art fusion methods integrate them at the feature level, but they mostly rely on the learned soft association between point clouds and images, which lacks interpretability and neglects the hard association between them. In this paper, we combine feature-level fusion with point-level fusion, using hard association established by the calibration matrices to guide the generation of object queries. Specifically, in the early fusion stage, we use the 2D CNN features of images to decorate the point cloud data, and employ two independent sparse convolutions to extract the decorated point cloud features. In the mid-level fusion stage, we initialize the queries with a center heatmap and embed the predicted class labels as auxiliary information into the queries, making the initial positions closer to the actual centers of the targets. Extensive experiments conducted on two popular datasets, i.e. KITTI, Waymo, demonstrate the superiority of DecoratingFusion.
Authors: Wei Chen, Yi Zhou
Abstract: In the realm of class-incremental learning (CIL), alleviating the catastrophic forgetting problem is a pivotal challenge. This paper discovers a counter-intuitive observation: by incorporating domain shift into CIL tasks, the forgetting rate is significantly reduced. Our comprehensive studies demonstrate that incorporating domain shift leads to a clearer separation in the feature distribution across tasks and helps reduce parameter interference during the learning process. Inspired by this observation, we propose a simple yet effective method named DisCo to deal with CIL tasks. DisCo introduces a lightweight prototype pool that utilizes contrastive learning to promote distinct feature distributions for the current task relative to previous ones, effectively mitigating interference across tasks. DisCo can be easily integrated into existing state-of-the-art class-incremental learning methods. Experimental results show that incorporating our method into various CIL methods achieves substantial performance improvements, validating the benefits of our approach in enhancing class-incremental learning by separating feature representation and reducing interference. These findings illustrate that DisCo can serve as a robust fashion for future research in class-incremental learning.
Authors: Edwin Arkel Rios, Jansen Christopher Yuanda, Vincent Leon Ghanz, Cheng-Wei Yu, Bo-Cheng Lai, Min-Chun Hu
Abstract: Ultra-fine-grained image recognition (UFGIR) is a challenging task that involves classifying images within a macro-category. While traditional FGIR deals with classifying different species, UFGIR goes beyond by classifying sub-categories within a species such as cultivars of a plant. In recent times the usage of Vision Transformer-based backbones has allowed methods to obtain outstanding recognition performances in this task but this comes at a significant cost in terms of computation specially since this task significantly benefits from incorporating higher resolution images. Therefore, techniques such as token reduction have emerged to reduce the computational cost. However, dropping tokens leads to loss of essential information for fine-grained categories, specially as the token keep rate is reduced. Therefore, to counteract the loss of information brought by the usage of token reduction we propose a novel Cross-Layer Aggregation Classification Head and a Cross-Layer Cache mechanism to recover and access information from previous layers in later locations. Extensive experiments covering more than 2000 runs across diverse settings including 5 datasets, 9 backbones, 7 token reduction methods, 5 keep rates, and 2 image sizes demonstrate the effectiveness of the proposed plug-and-play modules and allow us to push the boundaries of accuracy vs cost for UFGIR by reducing the kept tokens to extremely low ratios of up to 10\% while maintaining a competitive accuracy to state-of-the-art models. Code is available at: \url{https://github.com/arkel23/CLCA}
Authors: Zijie Li, Henry Li, Yichun Shi, Amir Barati Farimani, Yuval Kluger, Linjie Yang, Peng Wang
Abstract: Diffusion models have gained tremendous success in text-to-image generation, yet still lag behind with visual understanding tasks, an area dominated by autoregressive vision-language models. We propose a large-scale and fully end-to-end diffusion model for multi-modal understanding and generation that significantly improves on existing diffusion-based multimodal models, and is the first of its kind to support the full suite of vision-language modeling capabilities. Inspired by the multimodal diffusion transformer (MM-DiT) and recent advances in discrete diffusion language modeling, we leverage a cross-modal maximum likelihood estimation framework that simultaneously trains the conditional likelihoods of both images and text jointly under a single loss function, which is back-propagated through both branches of the diffusion transformer. The resulting model is highly flexible and capable of a wide range of tasks including image generation, captioning, and visual question answering. Our model attained competitive performance compared to recent unified image understanding and generation models, demonstrating the potential of multimodal diffusion modeling as a promising alternative to autoregressive next-token prediction models.
Authors: Haocheng Guo, Yaqiong Zhang, Lieyang Chen, Arfat Ahmad Khan
Abstract: The key to ensuring the safe obstacle avoidance function of autonomous driving systems lies in the use of extremely accurate vehicle recognition techniques. However, the variability of the actual road environment and the diverse characteristics of vehicles and pedestrians together constitute a huge obstacle to improving detection accuracy, posing a serious challenge to the realization of this goal. To address the above issues, this paper proposes an improved YOLOv8 vehicle detection method. Specifically, taking the YOLOv8n-seg model as the base model, firstly, the FasterNet network is used to replace the backbone network to achieve the purpose of reducing the computational complexity and memory while improving the detection accuracy and speed; secondly, the feature enhancement is achieved by adding the attention mechanism CBAM to the Neck; and lastly, the loss function CIoU is modified to WIoU, which optimizes the detection box localization while improving the segmentation accuracy. The results show that the improved model achieves 98.3%, 89.1% and 88.4% detection accuracy for car, Person and Motorcycle. Compared with the pre-improvement and YOLOv9 models in six metrics such as Precision.
Authors: Shi-Feng Peng, Guolei Sun, Yong Li, Hongsong Wang, Guo-Sen Xie
Abstract: The primary challenge of cross-domain few-shot segmentation (CD-FSS) is the domain disparity between the training and inference phases, which can exist in either the input data or the target classes. Previous models struggle to learn feature representations that generalize to various unknown domains from limited training domain samples. In contrast, the large-scale visual model SAM, pre-trained on tens of millions of images from various domains and classes, possesses excellent generalizability. In this work, we propose a SAM-aware graph prompt reasoning network (GPRN) that fully leverages SAM to guide CD-FSS feature representation learning and improve prediction accuracy. Specifically, we propose a SAM-aware prompt initialization module (SPI) to transform the masks generated by SAM into visual prompts enriched with high-level semantic information. Since SAM tends to divide an object into many sub-regions, this may lead to visual prompts representing the same semantic object having inconsistent or fragmented features. We further propose a graph prompt reasoning (GPR) module that constructs a graph among visual prompts to reason about their interrelationships and enable each visual prompt to aggregate information from similar prompts, thus achieving global semantic consistency. Subsequently, each visual prompt embeds its semantic information into the corresponding mask region to assist in feature representation learning. To refine the segmentation mask during testing, we also design a non-parameter adaptive point selection module (APS) to select representative point prompts from query predictions and feed them back to SAM to refine inaccurate segmentation results. Experiments on four standard CD-FSS datasets demonstrate that our method establishes new state-of-the-art results. Code: https://github.com/CVL-hub/GPRN.
Authors: Jiexin Wang, Yiju Guo, Bing Su
Abstract: Exploring the bridge between historical and future motion behaviors remains a central challenge in human motion prediction. While most existing methods incorporate a reconstruction task as an auxiliary task into the decoder, thereby improving the modeling of spatio-temporal dependencies, they overlook the potential conflicts between reconstruction and prediction tasks. In this paper, we propose a novel approach: Temporal Decoupling Decoding with Inverse Processing (\textbf{$TD^2IP$}). Our method strategically separates reconstruction and prediction decoding processes, employing distinct decoders to decode the shared motion features into historical or future sequences. Additionally, inverse processing reverses motion information in the temporal dimension and reintroduces it into the model, leveraging the bidirectional temporal correlation of human motion behaviors. By alleviating the conflicts between reconstruction and prediction tasks and enhancing the association of historical and future information, \textbf{$TD^2IP$} fosters a deeper understanding of motion patterns. Extensive experiments demonstrate the adaptability of our method within existing methods.
Authors: Jiexin Wang, Yiju Guo, Bing Su
Abstract: Human motion prediction (HMP) involves forecasting future human motion based on historical data. Graph Convolutional Networks (GCNs) have garnered widespread attention in this field for their proficiency in capturing relationships among joints in human motion. However, existing GCN-based methods tend to focus on either temporal-domain or spatial-domain features, or they combine spatio-temporal features without fully leveraging the complementarity and cross-dependency of these two features. In this paper, we propose the Spatial-Temporal Multi-Subgraph Graph Convolutional Network (STMS-GCN) to capture complex spatio-temporal dependencies in human motion. Specifically, we decouple the modeling of temporal and spatial dependencies, enabling cross-domain knowledge transfer at multiple scales through a spatio-temporal information consistency constraint mechanism. Besides, we utilize multiple subgraphs to extract richer motion information and enhance the learning associations of diverse subgraphs through a homogeneous information constraint mechanism. Extensive experiments on the standard HMP benchmarks demonstrate the superiority of our method.
Authors: Jicheol Park, Boseung Jeong, Dongwon Kim, Suha Kwak
Abstract: Text-based person search is the task of finding person images that are the most relevant to the natural language text description given as query. The main challenge of this task is a large gap between the target images and text queries, which makes it difficult to establish correspondence and distinguish subtle differences across people. To address this challenge, we introduce an efficient encoder-decoder model that extracts coarse-to-fine embedding vectors which are semantically aligned across the two modalities without supervision for the alignment. There is another challenge of learning to capture fine-grained information with only person IDs as supervision, where similar body parts of different individuals are considered different due to the lack of part-level supervision. To tackle this, we propose a novel ranking loss, dubbed commonality-based margin ranking loss, which quantifies the degree of commonality of each body part and reflects it during the learning of fine-grained body part details. As a consequence, it enables our method to achieve the best records on three public benchmarks.
Authors: Ling Fu, Biao Yang, Zhebin Kuang, Jiajun Song, Yuzhe Li, Linghao Zhu, Qidi Luo, Xinyu Wang, Hao Lu, Mingxin Huang, Zhang Li, Guozhi Tang, Bin Shan, Chunhui Lin, Qi Liu, Binghong Wu, Hao Feng, Hao Liu, Can Huang, Jingqun Tang, Wei Chen, Lianwen Jin, Yuliang Liu, Xiang Bai
Abstract: Scoring the Optical Character Recognition (OCR) capabilities of Large Multimodal Models (LMMs) has witnessed growing interest recently. Existing benchmarks have highlighted the impressive performance of LMMs in text recognition; however, their abilities on certain challenging tasks, such as text localization, handwritten content extraction, and logical reasoning, remain underexplored. To bridge this gap, we introduce OCRBench v2, a large-scale bilingual text-centric benchmark with currently the most comprehensive set of tasks (4x more tasks than the previous multi-scene benchmark OCRBench), the widest coverage of scenarios (31 diverse scenarios including street scene, receipt, formula, diagram, and so on), and thorough evaluation metrics, with a total of 10,000 human-verified question-answering pairs and a high proportion of difficult samples. After carefully benchmarking state-of-the-art LMMs on OCRBench v2, we find that 20 out of 22 LMMs score below 50 (100 in total) and suffer from five-type limitations, including less frequently encountered text recognition, fine-grained perception, layout perception, complex element parsing, and logical reasoning. The benchmark and evaluation scripts are available at https://github.com/Yuliang-liu/MultimodalOCR.
Authors: Runnan Chen, Xiangyu Sun, Zhaoqing Wang, Youquan Liu, Jiepeng Wang, Lingdong Kong, Jiankang Deng, Mingming Gong, Liang Pan, Wenping Wang, Tongliang Liu
Abstract: Open-vocabulary scene understanding using 3D Gaussian (3DGS) representations has garnered considerable attention. However, existing methods mostly lift knowledge from large 2D vision models into 3DGS on a scene-by-scene basis, restricting the capabilities of open-vocabulary querying within their training scenes so that lacking the generalizability to novel scenes. In this work, we propose \textbf{OVGaussian}, a generalizable \textbf{O}pen-\textbf{V}ocabulary 3D semantic segmentation framework based on the 3D \textbf{Gaussian} representation. We first construct a large-scale 3D scene dataset based on 3DGS, dubbed \textbf{SegGaussian}, which provides detailed semantic and instance annotations for both Gaussian points and multi-view images. To promote semantic generalization across scenes, we introduce Generalizable Semantic Rasterization (GSR), which leverages a 3D neural network to learn and predict the semantic property for each 3D Gaussian point, where the semantic property can be rendered as multi-view consistent 2D semantic maps. In the next, we propose a Cross-modal Consistency Learning (CCL) framework that utilizes open-vocabulary annotations of 2D images and 3D Gaussians within SegGaussian to train the 3D neural network capable of open-vocabulary semantic segmentation across Gaussian-based 3D scenes. Experimental results demonstrate that OVGaussian significantly outperforms baseline methods, exhibiting robust cross-scene, cross-domain, and novel-view generalization capabilities. Code and the SegGaussian dataset will be released. (https://github.com/runnanchen/OVGaussian).
Authors: Haifeng Zhao, Yuguang Jin, Leilei Ma
Abstract: Significant advancements have been made in single label incremental learning (SLCIL),yet the more practical and challenging multi label class incremental learning (MLCIL) remains understudied. Recently,visual language models such as CLIP have achieved good results in classification tasks. However,directly using CLIP to solve MLCIL issue can lead to catastrophic forgetting. To tackle this issue, we integrate an improved data replay mechanism and prompt loss to curb knowledge forgetting. Specifically,our model enhances the prompt information to better adapt to multi-label classification tasks and employs confidence-based replay strategy to select representative samples. Moreover, the prompt loss significantly reduces the model's forgetting of previous knowledge. Experimental results demonstrate that our method has substantially improved the performance of MLCIL tasks across multiple benchmark datasets,validating its effectiveness.
Authors: Yiwen Wang, Siyuan Chen, Ran Yi
Abstract: 3D Gaussian Splatting is emerging as a state-of-the-art technique in novel view synthesis, recognized for its impressive balance between visual quality, speed, and rendering efficiency. However, reliance on third-degree spherical harmonics for color representation introduces significant storage demands and computational overhead, resulting in a large memory footprint and slower rendering speed. We introduce SG-Splatting with Spherical Gaussians based color representation, a novel approach to enhance rendering speed and quality in novel view synthesis. Our method first represents view-dependent color using Spherical Gaussians, instead of three degree spherical harmonics, which largely reduces the number of parameters used for color representation, and significantly accelerates the rendering process. We then develop an efficient strategy for organizing multiple Spherical Gaussians, optimizing their arrangement to achieve a balanced and accurate scene representation. To further improve rendering quality, we propose a mixed representation that combines Spherical Gaussians with low-degree spherical harmonics, capturing both high- and low-frequency color information effectively. SG-Splatting also has plug-and-play capability, allowing it to be easily integrated into existing systems. This approach improves computational efficiency and overall visual fidelity, making it a practical solution for real-time applications.
Authors: Xiaolei Wang, Xiaoyang Wang, Huihui Bai, Eng Gee Lim, Jimin Xiao
Abstract: Existing unsupervised distillation-based methods rely on the differences between encoded and decoded features to locate abnormal regions in test images. However, the decoder trained only on normal samples still reconstructs abnormal patch features well, degrading performance. This issue is particularly pronounced in unsupervised multi-class anomaly detection tasks. We attribute this behavior to over-generalization(OG) of decoder: the significantly increasing diversity of patch patterns in multi-class training enhances the model generalization on normal patches, but also inadvertently broadens its generalization to abnormal patches. To mitigate OG, we propose a novel approach that leverages class-agnostic learnable prompts to capture common textual normality across various visual patterns, and then apply them to guide the decoded features towards a normal textual representation, suppressing over-generalization of the decoder on abnormal patterns. To further improve performance, we also introduce a gated mixture-of-experts module to specialize in handling diverse patch patterns and reduce mutual interference between them in multi-class training. Our method achieves competitive performance on the MVTec AD and VisA datasets, demonstrating its effectiveness.
Authors: Runnan Chen, Zhaoqing Wang, Jiepeng Wang, Yuexin Ma, Mingming Gong, Wenping Wang, Tongliang Liu
Abstract: Understanding geometric, semantic, and instance information in 3D scenes from sequential video data is essential for applications in robotics and augmented reality. However, existing Simultaneous Localization and Mapping (SLAM) methods generally focus on either geometric or semantic reconstruction. In this paper, we introduce PanoSLAM, the first SLAM system to integrate geometric reconstruction, 3D semantic segmentation, and 3D instance segmentation within a unified framework. Our approach builds upon 3D Gaussian Splatting, modified with several critical components to enable efficient rendering of depth, color, semantic, and instance information from arbitrary viewpoints. To achieve panoptic 3D scene reconstruction from sequential RGB-D videos, we propose an online Spatial-Temporal Lifting (STL) module that transfers 2D panoptic predictions from vision models into 3D Gaussian representations. This STL module addresses the challenges of label noise and inconsistencies in 2D predictions by refining the pseudo labels across multi-view inputs, creating a coherent 3D representation that enhances segmentation accuracy. Our experiments show that PanoSLAM outperforms recent semantic SLAM methods in both mapping and tracking accuracy. For the first time, it achieves panoptic 3D reconstruction of open-world environments directly from the RGB-D video. (https://github.com/runnanchen/PanoSLAM)
Authors: Yue Fan, Xiaojian Ma, Rongpeng Su, Jun Guo, Rujie Wu, Xi Chen, Qing Li
Abstract: This paper investigates the problem of understanding dynamic 3D scenes from egocentric observations, a key challenge in robotics and embodied AI. Unlike prior studies that explored this as long-form video understanding and utilized egocentric video only, we instead propose an LLM-based agent, Embodied VideoAgent, which constructs scene memory from both egocentric video and embodied sensory inputs (e.g. depth and pose sensing). We further introduce a VLM-based approach to automatically update the memory when actions or activities over objects are perceived. Embodied VideoAgent attains significant advantages over counterparts in challenging reasoning and planning tasks in 3D scenes, achieving gains of 4.9% on Ego4D-VQ3D, 5.8% on OpenEQA, and 11.7% on EnvQA. We have also demonstrated its potential in various embodied AI tasks including generating embodied interactions and perception for robot manipulation. The code and demo will be made public.
Authors: Dawen Yu, Shunping Ji
Abstract: Instance segmentation performance in remote sensing images (RSIs) is significantly affected by two issues: how to extract accurate boundaries of objects from remote imaging through the dynamic atmosphere, and how to integrate the mutual information of related object instances scattered over a vast spatial region. In this study, we propose a novel Shape Guided Transformer Network (SGTN) to accurately extract objects at the instance level. Inspired by the global contextual modeling capacity of the self-attention mechanism, we propose an effective transformer encoder termed LSwin, which incorporates vertical and horizontal 1D global self-attention mechanisms to obtain better global-perception capacity for RSIs than the popular local-shifted-window based Swin Transformer. To achieve accurate instance mask segmentation, we introduce a shape guidance module (SGM) to emphasize the object boundary and shape information. The combination of SGM, which emphasizes the local detail information, and LSwin, which focuses on the global context relationships, achieve excellent RSI instance segmentation. Their effectiveness was validated through comprehensive ablation experiments. Especially, LSwin is proved better than the popular ResNet and Swin transformer encoder at the same level of efficiency. Compared to other instance segmentation methods, our SGTN achieves the highest average precision (AP) scores on two single-class public datasets (WHU dataset and BITCC dataset) and a multi-class public dataset (NWPU VHR-10 dataset). Code will be available at http://gpcv.whu.edu.cn/data/.
Authors: Evelyn Zhang, Bang Xiao, Jiayi Tang, Qianli Ma, Chang Zou, Xuefei Ning, Xuming Hu, Linfeng Zhang
Abstract: Stable Diffusion has achieved remarkable success in the field of text-to-image generation, with its powerful generative capabilities and diverse generation results making a lasting impact. However, its iterative denoising introduces high computational costs and slows generation speed, limiting broader adoption. The community has made numerous efforts to reduce this computational burden, with methods like feature caching attracting attention due to their effectiveness and simplicity. Nonetheless, simply reusing features computed at previous timesteps causes the features across adjacent timesteps to become similar, reducing the dynamics of features over time and ultimately compromising the quality of generated images. In this paper, we introduce a dynamics-aware token pruning (DaTo) approach that addresses the limitations of feature caching. DaTo selectively prunes tokens with lower dynamics, allowing only high-dynamic tokens to participate in self-attention layers, thereby extending feature dynamics across timesteps. DaTo combines feature caching with token pruning in a training-free manner, achieving both temporal and token-wise information reuse. Applied to Stable Diffusion on the ImageNet, our approach delivered a 9$\times$ speedup while reducing FID by 0.33, indicating enhanced image quality. On the COCO-30k, we observed a 7$\times$ acceleration coupled with a notable FID reduction of 2.17.
Authors: Junmin Cai, Han Sun, Ningzhong Liu
Abstract: Camouflaged object detection (COD) aims to identify objects in images that are well hidden in the environment due to their high similarity to the background in terms of texture and color. However, existing most boundary-guided camouflage object detection algorithms tend to generate object boundaries early in the network, and inaccurate edge priors often introduce noises in object detection. Address on this issue, we propose a novel network named B2Net aiming to enhance the accuracy of obtained boundaries by reusing boundary-aware modules at different stages of the network. Specifically, we present a Residual Feature Enhanced Module (RFEM) with the goal of integrating more discriminative feature representations to enhance detection accuracy and reliability. After that, the Boundary Aware Module (BAM) is introduced to explore edge cues twice by integrating spatial information from low-level features and semantic information from high-level features. Finally, we design the Cross-scale Boundary Fusion Module(CBFM) that integrate information across different scales in a top-down manner, merging boundary features with object features to obtain a comprehensive feature representation incorporating boundary information. Extensive experimental results on three challenging benchmark datasets demonstrate that our proposed method B2Net outperforms 15 state-of-art methods under widely used evaluation metrics. Code will be made publicly available.
Authors: Lala Shakti Swarup Ray, Bo Zhou, Sungho Suh, Paul Lukowicz
Abstract: Understanding human-to-human interactions, especially in contexts like public security surveillance, is critical for monitoring and maintaining safety. Traditional activity recognition systems are limited by fixed vocabularies, predefined labels, and rigid interaction categories that often rely on choreographed videos and overlook concurrent interactive groups. These limitations make such systems less adaptable to real-world scenarios, where interactions are diverse and unpredictable. In this paper, we propose an open vocabulary human-to-human interaction recognition (OV-HHIR) framework that leverages large language models to generate open-ended textual descriptions of both seen and unseen human interactions in open-world settings without being confined to a fixed vocabulary. Additionally, we create a comprehensive, large-scale human-to-human interaction dataset by standardizing and combining existing public human interaction datasets into a unified benchmark. Extensive experiments demonstrate that our method outperforms traditional fixed-vocabulary classification systems and existing cross-modal language models for video understanding, setting the stage for more intelligent and adaptable visual understanding systems in surveillance and beyond.
Authors: Jianjie Luo, Jingwen Chen, Yehao Li, Yingwei Pan, Jianlin Feng, Hongyang Chao, Ting Yao
Abstract: Recently, zero-shot image captioning has gained increasing attention, where only text data is available for training. The remarkable progress in text-to-image diffusion model presents the potential to resolve this task by employing synthetic image-caption pairs generated by this pre-trained prior. Nonetheless, the defective details in the salient regions of the synthetic images introduce semantic misalignment between the synthetic image and text, leading to compromised results. To address this challenge, we propose a novel Patch-wise Cross-modal feature Mix-up (PCM) mechanism to adaptively mitigate the unfaithful contents in a fine-grained manner during training, which can be integrated into most of encoder-decoder frameworks, introducing our PCM-Net. Specifically, for each input image, salient visual concepts in the image are first detected considering the image-text similarity in CLIP space. Next, the patch-wise visual features of the input image are selectively fused with the textual features of the salient visual concepts, leading to a mixed-up feature map with less defective content. Finally, a visual-semantic encoder is exploited to refine the derived feature map, which is further incorporated into the sentence decoder for caption generation. Additionally, to facilitate the model training with synthetic data, a novel CLIP-weighted cross-entropy loss is devised to prioritize the high-quality image-text pairs over the low-quality counterparts. Extensive experiments on MSCOCO and Flickr30k datasets demonstrate the superiority of our PCM-Net compared with state-of-the-art VLMs-based approaches. It is noteworthy that our PCM-Net ranks first in both in-domain and cross-domain zero-shot image captioning. The synthetic dataset SynthImgCap and code are available at https://jianjieluo.github.io/SynthImgCap.
Authors: Yifan Xu, Xinhao Li, Yichun Yang, Rui Huang, Limin Wang
Abstract: The ability of perceiving fine-grained spatial and temporal information is crucial for video-language retrieval. However, the existing video retrieval benchmarks, such as MSRVTT and MSVD, fail to efficiently evaluate the fine-grained retrieval ability of video-language models (VLMs) due to a lack of detailed annotations. To address this problem, we present FIBER, a FIne-grained BEnchmark for text to video Retrieval, containing 1,000 videos sourced from the FineAction dataset. Uniquely, our FIBER benchmark provides detailed human-annotated spatial annotations and temporal annotations for each video, making it possible to independently evaluate the spatial and temporal bias of VLMs on video retrieval task. Besides, we employ a text embedding method to unlock the capability of fine-grained video-language understanding of Multimodal Large Language Models (MLLMs). Surprisingly, the experiment results show that our Video Large Language Encoder (VLLE) performs comparably to CLIP-based models on traditional benchmarks and has a stronger capability of fine-grained representation with lower spatial-temporal bias. Project page: https://fiber-bench.github.io.
Authors: Bao Q. Bui, Tien T. T. Nguyen, Duy M. Le, Cong Tran, Cuong Pham
Abstract: This paper presents a comprehensive study on the classification and detection of Silicosis-related lung inflammation. Our main contributions include 1) the creation of a newly curated chest X-ray (CXR) image dataset named SVBCX that is tailored to the nuances of lung inflammation caused by distinct agents, providing a valuable resource for silicosis and pneumonia research community; and 2) we propose a novel deep-learning architecture that integrates graph transformer networks alongside a traditional deep neural network module for the effective classification of silicosis and pneumonia. Additionally, we employ the Balanced Cross-Entropy (BalCE) as a loss function to ensure more uniform learning across different classes, enhancing the model's ability to discern subtle differences in lung conditions. The proposed model architecture and loss function selection aim to improve the accuracy and reliability of inflammation detection, particularly in the context of Silicosis. Furthermore, our research explores the efficacy of an ensemble approach that combines the strengths of diverse model architectures. Experimental results on the constructed dataset demonstrate promising outcomes, showcasing substantial enhancements compared to baseline models. The ensemble of models achieves a macro-F1 score of 0.9749 and AUC ROC scores exceeding 0.99 for each class, underscoring the effectiveness of our approach in accurate and robust lung inflammation classification.
Authors: Cheng Yuan, Jian Jiang, Kunyi Yang, Lv Wu, Rui Wang, Zi Meng, Haonan Ping, Ziyu Xu, Yifan Zhou, Wanli Song, Hesheng Wang, Qi Dou, Yutong Ban
Abstract: Surgery video segmentation is an important topic in the surgical AI field. It allows the AI model to understand the spatial information of a surgical scene. Meanwhile, due to the lack of annotated surgical data, surgery segmentation models suffer from limited performance. With the emergence of SAM2 model, a large foundation model for video segmentation trained on natural videos, zero-shot surgical video segmentation became more realistic but meanwhile remains to be explored. In this paper, we systematically evaluate the performance of SAM2 model in zero-shot surgery video segmentation task. We conducted experiments under different configurations, including different prompting strategies, robustness, etc. Moreover, we conducted an empirical evaluation over the performance, including 9 datasets with 17 different types of surgeries.
Authors: Madeleine Darbyshire, Elizabeth Sklar, Simon Parsons
Abstract: Precision agriculture leverages data and machine learning so that farmers can monitor their crops and target interventions precisely. This enables the precision application of herbicide only to weeds, or the precision application of fertilizer only to undernourished crops, rather than to the entire field. The approach promises to maximize yields while minimizing resource use and harm to the surrounding environment. To this end, we propose a hierarchical panoptic segmentation method that simultaneously determines leaf count (as an identifier of plant growth)and locates weeds within an image. In particular, our approach aims to improve the segmentation of smaller instances like the leaves and weeds by incorporating focal loss and boundary loss. Not only does this result in competitive performance, achieving a PQ+ of 81.89 on the standard training set, but we also demonstrate we can improve leaf-counting accuracy with our method. The code is available at https://github.com/madeleinedarbyshire/HierarchicalMask2Former.
URLs: https://github.com/madeleinedarbyshire/HierarchicalMask2Former.
Authors: Tiange Luo, Ang Cao, Gunhee Lee, Justin Johnson, Honglak Lee
Abstract: Despite recent advances in Vision-Language Models (VLMs), many still over-rely on visual language priors present in their training data rather than true visual reasoning. To examine the situation, we introduce ViLP, a visual question answering (VQA) benchmark that pairs each question with three potential answers and three corresponding images: one image whose answer can be inferred from text alone, and two images that demand visual reasoning. By leveraging image generative models, we ensure significant variation in texture, shape, conceptual combinations, hallucinated elements, and proverb-based contexts, making our benchmark images distinctly out-of-distribution. While humans achieve near-perfect accuracy, modern VLMs falter; for instance, GPT-4 achieves only 66.17% on ViLP. To alleviate this, we propose a self-improving framework in which models generate new VQA pairs and images, then apply pixel-level and semantic corruptions to form "good-bad" image pairs for self-training. Our training objectives compel VLMs to focus more on actual visual inputs and have demonstrated their effectiveness in enhancing the performance of open-source VLMs, including LLaVA-v1.5 and Cambrian.
Authors: Xinhao Li, Yi Wang, Jiashuo Yu, Xiangyu Zeng, Yuhan Zhu, Haian Huang, Jianfei Gao, Kunchang Li, Yinan He, Chenting Wang, Yu Qiao, Yali Wang, Limin Wang
Abstract: Long-context modeling is a critical capability for multimodal large language models (MLLMs), enabling them to process long-form contents with implicit memorization. Despite its advances, handling extremely long videos remains challenging due to the difficulty in maintaining crucial features over extended sequences. This paper introduces a Hierarchical visual token Compression (HiCo) method designed for high-fidelity representation and a practical context modeling system VideoChat-Flash tailored for multimodal long-sequence processing. HiCo capitalizes on the redundancy of visual information in long videos to compress long video context from the clip-level to the video-level, reducing the compute significantly while preserving essential details. VideoChat-Flash features a multi-stage short-to-long learning scheme, a rich dataset of real-world long videos named LongVid, and an upgraded "Needle-In-A-video-Haystack" (NIAH) for evaluating context capacities. In extensive experiments, VideoChat-Flash shows the leading performance on both mainstream long and short video benchmarks at the 7B model scale. It firstly gets 99.1% accuracy over 10,000 frames in NIAH among open-source models.
Authors: Zhenpeng Huang, Xinhao Li, Jiaqi Li, Jing Wang, Xiangyu Zeng, Cheng Liang, Tao Wu, Xi Chen, Liang Li, Limin Wang
Abstract: Multimodal Large Language Models (MLLMs) have shown significant progress in offline video understanding. However, applying these models to real-world scenarios, such as autonomous driving and human-computer interaction, presents unique challenges due to the need for real-time processing of continuous online video streams. To this end, this paper presents systematic efforts from three perspectives: evaluation benchmark, model architecture, and training strategy. First, we introduce OVBench, a comprehensive question-answering benchmark specifically designed to evaluate models' ability to perceive, memorize, and reason within online video contexts. It features six core task types across three temporal contexts-past, present, and future-forming 16 subtasks from diverse datasets. Second, we propose a new Pyramid Memory Bank (PMB) that effectively retains key spatiotemporal information in video streams. Third, we proposed an offline-to-online learning paradigm, designing an interleaved dialogue format for online video data and constructing an instruction-tuning dataset tailored for online video training. This framework led to the development of VideoChat-Online, a robust and efficient model for online video understanding. Despite the lower computational cost and higher efficiency, VideoChat-Online outperforms existing state-of-the-art offline and online models across popular offline video benchmarks and OVBench, demonstrating the effectiveness of our model architecture and training strategy.
Authors: Edgar Guzman, Robert D. Howe
Abstract: The unpredictable nature of outdoor settings introduces numerous safety concerns, making hazard detection crucial for safe navigation. This paper introduces a novel system for sidewalk safety navigation utilizing a hybrid approach that combines a Variational Autoencoder (VAE) with a One-Class Support Vector Machine (OCSVM). The system is designed to detect anomalies on sidewalks that could potentially pose walking hazards. A dataset comprising over 15,000 training frames and 5,000 testing frames was collected using video recordings, capturing various sidewalk scenarios, including normal and hazardous conditions. During deployment, the VAE utilizes its reconstruction mechanism to detect anomalies within a frame. Poor reconstruction by the VAE implies the presence of an anomaly, after which the OCSVM is used to confirm whether the anomaly is hazardous or non-hazardous. The proposed VAE model demonstrated strong performance, with a high Area Under the Curve (AUC) of 0.94, effectively distinguishing anomalies that could be potential hazards. The OCSVM is employed to reduce the detection of false hazard anomalies, such as manhole or water valve covers. This approach achieves an accuracy of 91.4%, providing a highly reliable system for distinguishing between hazardous and non-hazardous scenarios. These results suggest that the proposed system offers a robust solution for hazard detection in uncertain environments.
Authors: Yuqian Yuan, Hang Zhang, Wentong Li, Zesen Cheng, Boqiang Zhang, Long Li, Xin Li, Deli Zhao, Wenqiao Zhang, Yueting Zhuang, Jianke Zhu, Lidong Bing
Abstract: Video Large Language Models (Video LLMs) have recently exhibited remarkable capabilities in general video understanding. However, they mainly focus on holistic comprehension and struggle with capturing fine-grained spatial and temporal details. Besides, the lack of high-quality object-level video instruction data and a comprehensive benchmark further hinders their advancements. To tackle these challenges, we introduce the VideoRefer Suite to empower Video LLM for finer-level spatial-temporal video understanding, i.e., enabling perception and reasoning on any objects throughout the video. Specially, we thoroughly develop VideoRefer Suite across three essential aspects: dataset, model, and benchmark. Firstly, we introduce a multi-agent data engine to meticulously curate a large-scale, high-quality object-level video instruction dataset, termed VideoRefer-700K. Next, we present the VideoRefer model, which equips a versatile spatial-temporal object encoder to capture precise regional and sequential representations. Finally, we meticulously create a VideoRefer-Bench to comprehensively assess the spatial-temporal understanding capability of a Video LLM, evaluating it across various aspects. Extensive experiments and analyses demonstrate that our VideoRefer model not only achieves promising performance on video referring benchmarks but also facilitates general video understanding capabilities.
Authors: Jiageng Mao, Boyi Li, Boris Ivanovic, Yuxiao Chen, Yan Wang, Yurong You, Chaowei Xiao, Danfei Xu, Marco Pavone, Yue Wang
Abstract: Synthesizing photo-realistic visual observations from an ego vehicle's driving trajectory is a critical step towards scalable training of self-driving models. Reconstruction-based methods create 3D scenes from driving logs and synthesize geometry-consistent driving videos through neural rendering, but their dependence on costly object annotations limits their ability to generalize to in-the-wild driving scenarios. On the other hand, generative models can synthesize action-conditioned driving videos in a more generalizable way but often struggle with maintaining 3D visual consistency. In this paper, we present DreamDrive, a 4D spatial-temporal scene generation approach that combines the merits of generation and reconstruction, to synthesize generalizable 4D driving scenes and dynamic driving videos with 3D consistency. Specifically, we leverage the generative power of video diffusion models to synthesize a sequence of visual references and further elevate them to 4D with a novel hybrid Gaussian representation. Given a driving trajectory, we then render 3D-consistent driving videos via Gaussian splatting. The use of generative priors allows our method to produce high-quality 4D scenes from in-the-wild driving data, while neural rendering ensures 3D-consistent video generation from the 4D scenes. Extensive experiments on nuScenes and street view images demonstrate that DreamDrive can generate controllable and generalizable 4D driving scenes, synthesize novel views of driving videos with high fidelity and 3D consistency, decompose static and dynamic elements in a self-supervised manner, and enhance perception and planning tasks for autonomous driving.
Authors: Jiawei Yang, Jiahui Huang, Yuxiao Chen, Yan Wang, Boyi Li, Yurong You, Apoorva Sharma, Maximilian Igl, Peter Karkus, Danfei Xu, Boris Ivanovic, Yue Wang, Marco Pavone
Abstract: We present STORM, a spatio-temporal reconstruction model designed for reconstructing dynamic outdoor scenes from sparse observations. Existing dynamic reconstruction methods often rely on per-scene optimization, dense observations across space and time, and strong motion supervision, resulting in lengthy optimization times, limited generalization to novel views or scenes, and degenerated quality caused by noisy pseudo-labels for dynamics. To address these challenges, STORM leverages a data-driven Transformer architecture that directly infers dynamic 3D scene representations--parameterized by 3D Gaussians and their velocities--in a single forward pass. Our key design is to aggregate 3D Gaussians from all frames using self-supervised scene flows, transforming them to the target timestep to enable complete (i.e., "amodal") reconstructions from arbitrary viewpoints at any moment in time. As an emergent property, STORM automatically captures dynamic instances and generates high-quality masks using only reconstruction losses. Extensive experiments on public datasets show that STORM achieves precise dynamic scene reconstruction, surpassing state-of-the-art per-scene optimization methods (+4.3 to 6.6 PSNR) and existing feed-forward approaches (+2.1 to 4.7 PSNR) in dynamic regions. STORM reconstructs large-scale outdoor scenes in 200ms, supports real-time rendering, and outperforms competitors in scene flow estimation, improving 3D EPE by 0.422m and Acc5 by 28.02%. Beyond reconstruction, we showcase four additional applications of our model, illustrating the potential of self-supervised learning for broader dynamic scene understanding.
Authors: Yuchuan Tian, Jing Han, Chengcheng Wang, Yuchen Liang, Chao Xu, Hanting Chen
Abstract: Diffusion models have shown exceptional performance in visual generation tasks. Recently, these models have shifted from traditional U-Shaped CNN-Attention hybrid structures to fully transformer-based isotropic architectures. While these transformers exhibit strong scalability and performance, their reliance on complicated self-attention operation results in slow inference speeds. Contrary to these works, we rethink one of the simplest yet fastest module in deep learning, 3x3 Convolution, to construct a scaled-up purely convolutional diffusion model. We first discover that an Encoder-Decoder Hourglass design outperforms scalable isotropic architectures for Conv3x3, but still under-performing our expectation. Further improving the architecture, we introduce sparse skip connections to reduce redundancy and improve scalability. Based on the architecture, we introduce conditioning improvements including stage-specific embeddings, mid-block condition injection, and conditional gating. These improvements lead to our proposed Diffusion CNN (DiC), which serves as a swift yet competitive diffusion architecture baseline. Experiments on various scales and settings show that DiC surpasses existing diffusion transformers by considerable margins in terms of performance while keeping a good speed advantage. Project page: https://github.com/YuchuanTian/DiC
Authors: Sarah M. Hooper, Hui Xue
Abstract: Biomedical imaging modalities often produce high-resolution, multi-dimensional images that pose computational challenges for deep neural networks. These computational challenges are compounded when training transformers due to the self-attention operator, which scales quadratically with context length. Recent developments in long-context models have potential to alleviate these difficulties and enable more efficient application of transformers to large biomedical images, although a systematic evaluation on this topic is lacking. In this study, we investigate the impact of context length on biomedical image analysis and we evaluate the performance of recently proposed long-context models. We first curate a suite of biomedical imaging datasets, including 2D and 3D data for segmentation, denoising, and classification tasks. We then analyze the impact of context length on network performance using the Vision Transformer and Swin Transformer by varying patch size and attention window size. Our findings reveal a strong relationship between context length and performance, particularly for pixel-level prediction tasks. Finally, we show that recent long-context models demonstrate significant improvements in efficiency while maintaining comparable performance, though we highlight where gaps remain. This work underscores the potential and challenges of using long-context models in biomedical imaging.
Authors: Kyle Gao, Liangzhi Li, Hongjie He, Dening Lu, Linlin Xu, Jonathan Li
Abstract: Recently released open-source pre-trained foundational image segmentation and object detection models (SAM2+GroundingDINO) allow for geometrically consistent segmentation of objects of interest in multi-view 2D images. Users can use text-based or click-based prompts to segment objects of interest without requiring labeled training datasets. Gaussian Splatting allows for the learning of the 3D representation of a scene's geometry and radiance based on 2D images. Combining Google Earth Studio, SAM2+GroundingDINO, 2D Gaussian Splatting, and our improvements in mask refinement based on morphological operations and contour simplification, we created a pipeline to extract the 3D mesh of any building based on its name, address, or geographic coordinates.
Authors: Tianfu Wang, Mingyang Xie, Haoming Cai, Sachin Shah, Christopher A. Metzler
Abstract: Transparent surfaces, such as glass, create complex reflections that obscure images and challenge downstream computer vision applications. We introduce Flash-Split, a robust framework for separating transmitted and reflected light using a single (potentially misaligned) pair of flash/no-flash images. Our core idea is to perform latent-space reflection separation while leveraging the flash cues. Specifically, Flash-Split consists of two stages. Stage 1 separates apart the reflection latent and transmission latent via a dual-branch diffusion model conditioned on an encoded flash/no-flash latent pair, effectively mitigating the flash/no-flash misalignment issue. Stage 2 restores high-resolution, faithful details to the separated latents, via a cross-latent decoding process conditioned on the original images before separation. By validating Flash-Split on challenging real-world scenes, we demonstrate state-of-the-art reflection separation performance and significantly outperform the baseline methods.
Authors: Kim Sung-Bin, Kim Jun-Seong, Junseok Ko, Yewon Kim, Tae-Hyun Oh
Abstract: We propose SoundBrush, a model that uses sound as a brush to edit and manipulate visual scenes. We extend the generative capabilities of the Latent Diffusion Model (LDM) to incorporate audio information for editing visual scenes. Inspired by existing image-editing works, we frame this task as a supervised learning problem and leverage various off-the-shelf models to construct a sound-paired visual scene dataset for training. This richly generated dataset enables SoundBrush to learn to map audio features into the textual space of the LDM, allowing for visual scene editing guided by diverse in-the-wild sound. Unlike existing methods, SoundBrush can accurately manipulate the overall scenery or even insert sounding objects to best match the audio inputs while preserving the original content. Furthermore, by integrating with novel view synthesis techniques, our framework can be extended to edit 3D scenes, facilitating sound-driven 3D scene manipulation. Demos are available at https://soundbrush.github.io/.
Authors: Suttisak Wizadwongsa, Jinfan Zhou, Edward Li, Jeong Joon Park
Abstract: Recent AI-based 3D content creation has largely evolved along two paths: feed-forward image-to-3D reconstruction approaches and 3D generative models trained with 2D or 3D supervision. In this work, we show that existing feed-forward reconstruction methods can serve as effective latent encoders for training 3D generative models, thereby bridging these two paradigms. By reusing powerful pre-trained reconstruction models, we avoid computationally expensive encoder network training and obtain rich 3D latent features for generative modeling for free. However, the latent spaces of reconstruction models are not well-suited for generative modeling due to their unstructured nature. To enable flow-based model training on these latent features, we develop post-processing pipelines, including protocols to standardize the features and spatial weighting to concentrate on important regions. We further incorporate a 2D image space perceptual rendering loss to handle the high-dimensional latent spaces. Finally, we propose a multi-stream transformer-based rectified flow architecture to achieve linear scaling and high-quality text-conditioned 3D generation. Our framework leverages the advancements of feed-forward reconstruction models to enhance the scalability of 3D generative modeling, achieving both high computational efficiency and state-of-the-art performance in text-to-3D generation.
Authors: Xindi Wu, Mengzhou Xia, Rulin Shao, Zhiwei Deng, Pang Wei Koh, Olga Russakovsky
Abstract: Visual Instruction Tuning typically requires a large amount of vision-language training data. This data often containing redundant information that increases computational costs without proportional performance gains. In this work, we introduce ICONS, a gradient-driven Influence CONsensus approach for vision-language data Selection that selects a compact training dataset for efficient multi-task training. The key element of our approach is cross-task influence consensus, which uses majority voting across task-specific influence matrices to identify samples that are consistently valuable across multiple tasks, allowing us to effectively prioritize data that optimizes for overall performance. Experiments show that models trained on our selected data (20% of LLaVA-665K) achieve 98.6% of the relative performance obtained using the full dataset. Additionally, we release this subset, LLaVA-ICONS-133K, a compact yet highly informative subset of LLaVA-665K visual instruction tuning data, preserving high impact training data for efficient vision-language model development.
Authors: Hao Wang, Cheng Deng, Zhidong Zhao
Abstract: Recent generative models demonstrate impressive performance on synthesizing photographic images, which makes humans hardly to distinguish them from pristine ones, especially on realistic-looking synthetic facial images. Previous works mostly focus on mining discriminative artifacts from vast amount of visual data. However, they usually lack the exploration of prior knowledge and rarely pay attention to the domain shift between training categories (e.g., natural and indoor objects) and testing ones (e.g., fine-grained human facial images), resulting in unsatisfactory detection performance. To address these issues, we propose a novel knowledge-guided prompt learning method for deepfake facial image detection. Specifically, we retrieve forgery-related prompts from large language models as expert knowledge to guide the optimization of learnable prompts. Besides, we elaborate test-time prompt tuning to alleviate the domain shift, achieving significant performance improvement and boosting the application in real-world scenarios. Extensive experiments on DeepFakeFaceForensics dataset show that our proposed approach notably outperforms state-of-the-art methods.
Authors: Hui Zeng, Sanshuai Cui, Biwei Chen, Anjie Peng
Abstract: Adversarial examples' (AE) transferability refers to the phenomenon that AEs crafted with one surrogate model can also fool other models. Notwithstanding remarkable progress in untargeted transferability, its targeted counterpart remains challenging. This paper proposes an everywhere scheme to boost targeted transferability. Our idea is to attack a victim image both globally and locally. We aim to optimize 'an army of targets' in every local image region instead of the previous works that optimize a high-confidence target in the image. Specifically, we split a victim image into non-overlap blocks and jointly mount a targeted attack on each block. Such a strategy mitigates transfer failures caused by attention inconsistency between surrogate and victim models and thus results in stronger transferability. Our approach is method-agnostic, which means it can be easily combined with existing transferable attacks for even higher transferability. Extensive experiments on ImageNet demonstrate that the proposed approach universally improves the state-of-the-art targeted attacks by a clear margin, e.g., the transferability of the widely adopted Logit attack can be improved by 28.8%-300%.We also evaluate the crafted AEs on a real-world platform: Google Cloud Vision. Results further support the superiority of the proposed method.
Authors: Yuji Arima, Satoshi Kagiwada, Hitoshi Iyatomi
Abstract: Recent studies on plant disease diagnosis using machine learning (ML) have highlighted concerns about the overestimated diagnostic performance due to inappropriate data partitioning, where training and test datasets are derived from the same source (domain). Plant disease diagnosis presents a challenging classification task, characterized by its fine-grained nature, vague symptoms, and the extensive variability of image features within each domain. In this study, we propose the concept of Discriminative Difficulty Distance (DDD), a novel metric designed to quantify the domain gap between training and test datasets while assessing the classification difficulty of test data. DDD provides a valuable tool for identifying insufficient diversity in training data, thus supporting the development of more diverse and robust datasets. We investigated multiple image encoders trained on different datasets and examined whether the distances between datasets, measured using low-dimensional representations generated by the encoders, are suitable as a DDD metric. The study utilized 244,063 plant disease images spanning four crops and 34 disease classes collected from 27 domains. As a result, we demonstrated that even if the test images are from different crops or diseases than those used to train the encoder, incorporating them allows the construction of a distance measure for a dataset that strongly correlates with the difficulty of diagnosis indicated by the disease classifier developed independently. Compared to the base encoder, pre-trained only on ImageNet21K, the correlation higher by 0.106 to 0.485, reaching a maximum of 0.909.
Authors: Ruibin Li, Tao Yang, Song Guo, Lei Zhang
Abstract: Despite the significant advancements, existing object removal methods struggle with incomplete removal, incorrect content synthesis and blurry synthesized regions, resulting in low success rates. Such issues are mainly caused by the lack of high-quality paired training data, as well as the self-supervised training paradigm adopted in these methods, which forces the model to in-paint the masked regions, leading to ambiguity between synthesizing the masked objects and restoring the background. To address these issues, we propose a semi-supervised learning strategy with human-in-the-loop to create high-quality paired training data, aiming to train a Robust Object Remover (RORem). We first collect 60K training pairs from open-source datasets to train an initial object removal model for generating removal samples, and then utilize human feedback to select a set of high-quality object removal pairs, with which we train a discriminator to automate the following training data generation process. By iterating this process for several rounds, we finally obtain a substantial object removal dataset with over 200K pairs. Fine-tuning the pre-trained stable diffusion model with this dataset, we obtain our RORem, which demonstrates state-of-the-art object removal performance in terms of both reliability and image quality. Particularly, RORem improves the object removal success rate over previous methods by more than 18\%. The dataset, source code and trained model are available at https://github.com/leeruibin/RORem.
Authors: Chuanzhi Xu, Langyi Chen, Vincent Qu, Haodong Chen, Vera Chung
Abstract: Neuromorphic cameras, also known as event cameras, are asynchronous brightness-change sensors that can capture extremely fast motion without suffering from motion blur, making them particularly promising for 3D reconstruction in extreme environments. However, existing research on 3D reconstruction using monocular neuromorphic cameras is limited, and most of the methods rely on estimating physical priors and employ complex multi-step pipelines. In this work, we propose an end-to-end method for dense voxel 3D reconstruction using neuromorphic cameras that eliminates the need to estimate physical priors. Our method incorporates a novel event representation to enhance edge features, enabling the proposed feature-enhancement model to learn more effectively. Additionally, we introduced Optimal Binarization Threshold Selection Principle as a guideline for future related work, using the optimal reconstruction results achieved with threshold optimization as the benchmark. Our method achieves a 54.6% improvement in reconstruction accuracy compared to the baseline method.
Authors: Suho Park, SuBeen Lee, Hyun Seok Seong, Jaejoon Yoo, Jae-Pil Heo
Abstract: We propose Foreground-Covering Prototype Generation and Matching to resolve Few-Shot Segmentation (FSS), which aims to segment target regions in unlabeled query images based on labeled support images. Unlike previous research, which typically estimates target regions in the query using support prototypes and query pixels, we utilize the relationship between support and query prototypes. To achieve this, we utilize two complementary features: SAM Image Encoder features for pixel aggregation and ResNet features for class consistency. Specifically, we construct support and query prototypes with SAM features and distinguish query prototypes of target regions based on ResNet features. For the query prototype construction, we begin by roughly guiding foreground regions within SAM features using the conventional pseudo-mask, then employ iterative cross-attention to aggregate foreground features into learnable tokens. Here, we discover that the cross-attention weights can effectively alternate the conventional pseudo-mask. Therefore, we use the attention-based pseudo-mask to guide ResNet features to focus on the foreground, then infuse the guided ResNet feature into the learnable tokens to generate class-consistent query prototypes. The generation of the support prototype is conducted symmetrically to that of the query one, with the pseudo-mask replaced by the ground-truth mask. Finally, we compare these query prototypes with support ones to generate prompts, which subsequently produce object masks through the SAM Mask Decoder. Our state-of-the-art performances on various datasets validate the effectiveness of the proposed method for FSS. Our official code is available at https://github.com/SuhoPark0706/FCP
Authors: Chenlong Xu, Bineng Zhong, Qihua Liang, Yaozong Zheng, Guorong Li, Shuxiang Song
Abstract: Recently, several studies have shown that utilizing contextual information to perceive target states is crucial for object tracking. They typically capture context by incorporating multiple video frames. However, these naive frame-context methods fail to consider the importance of each patch within a reference frame, making them susceptible to noise and redundant tokens, which deteriorates tracking performance. To address this challenge, we propose a new token context-aware tracking pipeline named LMTrack, designed to automatically learn high-quality reference tokens for efficient visual tracking. Embracing the principle of Less is More, the core idea of LMTrack is to analyze the importance distribution of all reference tokens, where important tokens are collected, continually attended to, and updated. Specifically, a novel Token Context Memory module is designed to dynamically collect high-quality spatio-temporal information of a target in an autoregressive manner, eliminating redundant background tokens from the reference frames. Furthermore, an effective Unidirectional Token Attention mechanism is designed to establish dependencies between reference tokens and search frame, enabling robust cross-frame association and target localization. Extensive experiments demonstrate the superiority of our tracker, achieving state-of-the-art results on tracking benchmarks such as GOT-10K, TrackingNet, and LaSOT.
Authors: Yulong Li, Yuxuan Zhang, Feilong Tang, Mian Zhou, Zhixiang Lu, Haochen Xue, Yifang Wang, Kang Dang, Jionglong Su
Abstract: Although sign language recognition aids non-hearing-impaired understanding, many hearing-impaired individuals still rely on sign language alone due to limited literacy, underscoring the need for advanced sign language production and translation (SLP and SLT) systems. In the field of sign language production, the lack of adequate models and datasets restricts practical applications. Existing models face challenges in production accuracy and pose control, making it difficult to provide fluent sign language expressions across diverse scenarios. Additionally, data resources are scarce, particularly high-quality datasets with complete sign vocabulary and pose annotations. To address these issues, we introduce CNText2Sign and CNSign, comprehensive datasets to benchmark SLP and SLT, respectively, with CNText2Sign covering gloss and landmark mappings for SLP, and CNSign providing extensive video-to-text data for SLT. To improve the accuracy and applicability of sign language systems, we propose the AuraLLM and SignMST-C models. AuraLLM, incorporating LoRA and RAG techniques, achieves a BLEU-4 score of 50.41 on the CNText2Sign dataset, enabling precise control over gesture semantics and motion. SignMST-C employs self-supervised rapid motion video pretraining, achieving a BLEU-4 score of 31.03/32.08 on the PHOENIX2014-T benchmark, setting a new state-of-the-art. These models establish robust baselines for the datasets released for their respective tasks.
Authors: Binglu Wang, Yao Tian, Shunzhou Wang, Le Yang
Abstract: The task of long-term action anticipation demands solutions that can effectively model temporal dynamics over extended periods while deeply understanding the inherent semantics of actions. Traditional approaches, which primarily rely on recurrent units or Transformer layers to capture long-term dependencies, often fall short in addressing these challenges. Large Language Models (LLMs), with their robust sequential modeling capabilities and extensive commonsense knowledge, present new opportunities for long-term action anticipation. In this work, we introduce the ActionLLM framework, a novel approach that treats video sequences as successive tokens, leveraging LLMs to anticipate future actions. Our baseline model simplifies the LLM architecture by setting future tokens, incorporating an action tuning module, and reducing the textual decoder layer to a linear layer, enabling straightforward action prediction without the need for complex instructions or redundant descriptions. To further harness the commonsense reasoning of LLMs, we predict action categories for observed frames and use sequential textual clues to guide semantic understanding. In addition, we introduce a Cross-Modality Interaction Block, designed to explore the specificity within each modality and capture interactions between vision and textual modalities, thereby enhancing multimodal tuning. Extensive experiments on benchmark datasets demonstrate the superiority of the proposed ActionLLM framework, encouraging a promising direction to explore LLMs in the context of action anticipation. Code is available at https://github.com/2tianyao1/ActionLLM.git.
Authors: Erik Nguyen, Spencer Htin
Abstract: The use of beauty filters on social media, which enhance the appearance of individuals in images, is a well-researched area, with existing methods proving to be highly effective. Traditionally, such enhancements are performed using rule-based approaches that leverage domain knowledge of facial features associated with attractiveness, applying very specific transformations to maximize these attributes. In this work, we present an alternative approach that projects facial images as points on the latent space of a pre-trained GAN, which are then optimized to produce beautiful faces. The movement of the latent points is guided by a newly developed facial beauty evaluation regression network, which learns to distinguish attractive facial features, outperforming many existing facial beauty evaluation models in this domain. By using this data-driven approach, our method can automatically capture holistic patterns in beauty directly from data rather than relying on predefined rules, enabling more dynamic and potentially broader applications of facial beauty editing. This work demonstrates a potential new direction for automated aesthetic enhancement, offering a complementary alternative to existing methods.
Authors: Rui Yang, Xiaojun Wu, Shengfeng He
Abstract: Current sketch extraction methods either require extensive training or fail to capture a wide range of artistic styles, limiting their practical applicability and versatility. We introduce Mixture-of-Self-Attention (MixSA), a training-free sketch extraction method that leverages strong diffusion priors for enhanced sketch perception. At its core, MixSA employs a mixture-of-self-attention technique, which manipulates self-attention layers by substituting the keys and values with those from reference sketches. This allows for the seamless integration of brushstroke elements into initial outline images, offering precise control over texture density and enabling interpolation between styles to create novel, unseen styles. By aligning brushstroke styles with the texture and contours of colored images, particularly in late decoder layers handling local textures, MixSA addresses the common issue of color averaging by adjusting initial outlines. Evaluated with various perceptual metrics, MixSA demonstrates superior performance in sketch quality, flexibility, and applicability. This approach not only overcomes the limitations of existing methods but also empowers users to generate diverse, high-fidelity sketches that more accurately reflect a wide range of artistic expressions.
Authors: Xinru Meng, Han Sun, Jiamei Liu, Ningzhong Liu, Huiyu Zhou
Abstract: Test-time Adaptation (TTA) aims to improve model performance when the model encounters domain changes after deployment. The standard TTA mainly considers the case where the target domain is static, while the continual TTA needs to undergo a sequence of domain changes. This encounters a significant challenge as the model needs to adapt for the long-term and is unaware of when the domain changes occur. The quality of pseudo-labels is hard to guarantee. Noisy pseudo-labels produced by simple self-training methods can cause error accumulation and catastrophic forgetting. In this work, we propose a new framework named SPARNet which consists of two parts, sample partitioning strategy and anti-forgetting regularization. The sample partition strategy divides samples into two groups, namely reliable samples and unreliable samples. According to the characteristics of each group of samples, we choose different strategies to deal with different groups of samples. This ensures that reliable samples contribute more to the model. At the same time, the negative impacts of unreliable samples are eliminated by the mean teacher's consistency learning. Finally, we introduce a regularization term to alleviate the catastrophic forgetting problem, which can limit important parameters from excessive changes. This term enables long-term adaptation of parameters in the network. The effectiveness of our method is demonstrated in continual TTA scenario by conducting a large number of experiments on CIFAR10-C, CIFAR100-C and ImageNet-C.
Authors: Gur Elkin, Ofir Itzhak Shahar, Yaniv Ohayon, Nadav Alali, Ohad Ben-Shahar
Abstract: Ancient artworks obtained in archaeological excavations usually suffer from a certain degree of fragmentation and physical degradation. Often, fragments of multiple artifacts from different periods or artistic styles could be found on the same site. With each fragment containing only partial information about its source, and pieces from different objects being mixed, categorizing broken artifacts based on their visual cues could be a challenging task, even for professionals. As classification is a common function of many machine learning models, the power of modern architectures can be harnessed for efficient and accurate fragment classification. In this work, we present a generalized deep-learning framework for predicting the artistic style of image fragments, achieving state-of-the-art results for pieces with varying styles and geometries.
Authors: Qianang Zhou, Junhui Hou, Meiyi Yang, Yongjian Deng, Youfu Li, Junlin Xiong
Abstract: Current optical flow methods exploit the stable appearance of frame (or RGB) data to establish robust correspondences across time. Event cameras, on the other hand, provide high-temporal-resolution motion cues and excel in challenging scenarios. These complementary characteristics underscore the potential of integrating frame and event data for optical flow estimation. However, most cross-modal approaches fail to fully utilize the complementary advantages, relying instead on simply stacking information. This study introduces a novel approach that uses a spatially dense modality to guide the aggregation of the temporally dense event modality, achieving effective cross-modal fusion. Specifically, we propose an event-enhanced frame representation that preserves the rich texture of frames and the basic structure of events. We use the enhanced representation as the guiding modality and employ events to capture temporally dense motion information. The robust motion features derived from the guiding modality direct the aggregation of motion information from events. To further enhance fusion, we propose a transformer-based module that complements sparse event motion features with spatially rich frame information and enhances global information propagation. Additionally, a mix-fusion encoder is designed to extract comprehensive spatiotemporal contextual features from both modalities. Extensive experiments on the MVSEC and DSEC-Flow datasets demonstrate the effectiveness of our framework. Leveraging the complementary strengths of frames and events, our method achieves leading performance on the DSEC-Flow dataset. Compared to the event-only model, frame guidance improves accuracy by 10\%. Furthermore, it outperforms the state-of-the-art fusion-based method with a 4\% accuracy gain and a 45\% reduction in inference time.
Authors: Nathanael L. Baisa
Abstract: In this work, we investigate four different fusion methods for associating detections to tracklets in multi-object visual tracking. In addition to considering strong cues such as motion and appearance information, we also consider weak cues such as height intersection-over-union (height-IoU) and tracklet confidence information in the data association using different fusion methods. These fusion methods include minimum, weighted sum based on IoU, Kalman filter (KF) gating, and hadamard product of costs due to the different cues. We conduct extensive evaluations on validation sets of MOT17, MOT20 and DanceTrack datasets, and find out that the choice of a fusion method is key for data association in multi-object visual tracking. We hope that this investigative work helps the computer vision research community to use the right fusion method for data association in multi-object visual tracking.
Authors: Yiming Zhang, Zicheng Zhang, Xinyi Wei, Xiaohong Liu, Guangtao Zhai, Xiongkuo Min
Abstract: Current Visual Language Models (VLMs) show impressive image understanding but struggle with visual illusions, especially in real-world scenarios. Existing benchmarks focus on classical cognitive illusions, which have been learned by state-of-the-art (SOTA) VLMs, revealing issues such as hallucinations and limited perceptual abilities. To address this gap, we introduce IllusionBench, a comprehensive visual illusion dataset that encompasses not only classic cognitive illusions but also real-world scene illusions. This dataset features 1,051 images, 5,548 question-answer pairs, and 1,051 golden text descriptions that address the presence, causes, and content of the illusions. We evaluate ten SOTA VLMs on this dataset using true-or-false, multiple-choice, and open-ended tasks. In addition to real-world illusions, we design trap illusions that resemble classical patterns but differ in reality, highlighting hallucination issues in SOTA models. The top-performing model, GPT-4o, achieves 80.59% accuracy on true-or-false tasks and 76.75% on multiple-choice questions, but still lags behind human performance. In the semantic description task, GPT-4o's hallucinations on classical illusions result in low scores for trap illusions, even falling behind some open-source models. IllusionBench is, to the best of our knowledge, the largest and most comprehensive benchmark for visual illusions in VLMs to date.
Authors: Kun Li, George Vosselman, Michael Ying Yang
Abstract: The goal of referring remote sensing image segmentation (RRSIS) is to extract specific pixel-level regions within an aerial image via a natural language expression. Recent advancements, particularly Transformer-based fusion designs, have demonstrated remarkable progress in this domain. However, existing methods primarily focus on refining visual features using language-aware guidance during the cross-modal fusion stage, neglecting the complementary vision-to-language flow. This limitation often leads to irrelevant or suboptimal representations. In addition, the diverse spatial scales of ground objects in aerial images pose significant challenges to the visual perception capabilities of existing models when conditioned on textual inputs. In this paper, we propose an innovative framework called Scale-wise Bidirectional Alignment Network (SBANet) to address these challenges for RRSIS. Specifically, we design a Bidirectional Alignment Module (BAM) with learnable query tokens to selectively and effectively represent visual and linguistic features, emphasizing regions associated with key tokens. BAM is further enhanced with a dynamic feature selection block, designed to provide both macro- and micro-level visual features, preserving global context and local details to facilitate more effective cross-modal interaction. Furthermore, SBANet incorporates a text-conditioned channel and spatial aggregator to bridge the gap between the encoder and decoder, enhancing cross-scale information exchange in complex aerial scenarios. Extensive experiments demonstrate that our proposed method achieves superior performance in comparison to previous state-of-the-art methods on the RRSIS-D and RefSegRS datasets, both quantitatively and qualitatively. The code will be released after publication.
Authors: Mingjia Li, Shuang Li, Tongrui Su, Longhui Yuan, Jian Liang, Wei Li
Abstract: Capitalizing on the complementary advantages of generative and discriminative models has always been a compelling vision in machine learning, backed by a growing body of research. This work discloses the hidden semantic structure within score-based generative models, unveiling their potential as effective discriminative priors. Inspired by our theoretical findings, we propose DUSA to exploit the structured semantic priors underlying diffusion score to facilitate the test-time adaptation of image classifiers or dense predictors. Notably, DUSA extracts knowledge from a single timestep of denoising diffusion, lifting the curse of Monte Carlo-based likelihood estimation over timesteps. We demonstrate the efficacy of our DUSA in adapting a wide variety of competitive pre-trained discriminative models on diverse test-time scenarios. Additionally, a thorough ablation study is conducted to dissect the pivotal elements in DUSA. Code is publicly available at https://github.com/BIT-DA/DUSA.
Authors: Bingyu Li, Da Zhang, Zhiyuan Zhao, Junyu Gao, Xuelong Li
Abstract: Open-vocabulary segmentation aims to identify and segment specific regions and objects based on text-based descriptions. A common solution is to leverage powerful vision-language models (VLMs), such as CLIP, to bridge the gap between vision and text information. However, VLMs are typically pretrained for image-level vision-text alignment, focusing on global semantic features. In contrast, segmentation tasks require fine-grained pixel-level alignment and detailed category boundary information, which VLMs alone cannot provide. As a result, information extracted directly from VLMs can't meet the requirements of segmentation tasks. To address this limitation, we propose FGAseg, a model designed for fine-grained pixel-text alignment and category boundary supplementation. The core of FGAseg is a Pixel-Level Alignment module that employs a cross-modal attention mechanism and a text-pixel alignment loss to refine the coarse-grained alignment from CLIP, achieving finer-grained pixel-text semantic alignment. Additionally, to enrich category boundary information, we introduce the alignment matrices as optimizable pseudo-masks during forward propagation and propose Category Information Supplementation module. These pseudo-masks, derived from cosine and convolutional similarity, provide essential global and local boundary information between different categories. By combining these two strategies, FGAseg effectively enhances pixel-level alignment and category boundary information, addressing key challenges in open-vocabulary segmentation. Extensive experiments demonstrate that FGAseg outperforms existing methods on open-vocabulary semantic segmentation benchmarks.
Authors: Teng Hu, Jiangning Zhang, Ran Yi, Jieyu Weng, Yabiao Wang, Xianfang Zeng, Zhucun Xue, Lizhuang Ma
Abstract: Employing LLMs for visual generation has recently become a research focus. However, the existing methods primarily transfer the LLM architecture to visual generation but rarely investigate the fundamental differences between language and vision. This oversight may lead to suboptimal utilization of visual generation capabilities within the LLM framework. In this paper, we explore the characteristics of visual embedding space under the LLM framework and discover that the correlation between visual embeddings can help achieve more stable and robust generation results. We present IAR, an Improved AutoRegressive Visual Generation Method that enhances the training efficiency and generation quality of LLM-based visual generation models. Firstly, we propose a Codebook Rearrangement strategy that uses balanced k-means clustering algorithm to rearrange the visual codebook into clusters, ensuring high similarity among visual features within each cluster. Leveraging the rearranged codebook, we propose a Cluster-oriented Cross-entropy Loss that guides the model to correctly predict the cluster where the token is located. This approach ensures that even if the model predicts the wrong token index, there is a high probability the predicted token is located in the correct cluster, which significantly enhances the generation quality and robustness. Extensive experiments demonstrate that our method consistently enhances the model training efficiency and performance from 100M to 1.4B, reducing the training time by half while achieving the same FID. Additionally, our approach can be applied to various LLM-based visual generation models and adheres to the scaling law, providing a promising direction for future research in LLM-based visual generation.
Authors: Libin Lan, Lu Jiang, Tianshu Yu, Xiaojuan Liu, Zhongshi He
Abstract: Video summarization mainly aims to produce a compact, short, informative, and representative synopsis of raw videos, which is of great importance for browsing, analyzing, and understanding video content. Dominant video summarization approaches are generally based on recurrent or convolutional neural networks, even recent encoder-only transformers. We propose using full transformer as an alternative architecture to perform video summarization. The full transformer with an encoder-decoder structure, specifically designed for handling sequence transduction problems, is naturally suitable for video summarization tasks. This work considers supervised video summarization and casts it as a sequence-to-sequence learning problem. Our key idea is to directly apply the full transformer to the video summarization task, which is intuitively sound and effective. Also, considering the efficiency problem, we replace full attention with the combination of local and global sparse attention, which enables modeling long-range dependencies while reducing computational costs. Based on this, we propose a transformer-like architecture, named FullTransNet, which has a full encoder-decoder structure with local-global sparse attention for video summarization. Specifically, both the encoder and decoder in FullTransNet are stacked the same way as ones in the vanilla transformer, and the local-global sparse attention is used only at the encoder side. Extensive experiments on two public multimedia benchmark datasets SumMe and TVSum demonstrate that our proposed model can outperform other video summarization approaches, achieving F-Measures of 54.4% on SumMe and 63.9% on TVSum with relatively lower compute and memory requirements, verifying its effectiveness and efficiency. The code and models are publicly available on GitHub.
Authors: Chenyang Liu, Keyan Chen, Rui Zhao, Zhengxia Zou, Zhenwei Shi
Abstract: Generative foundation models have advanced large-scale text-driven natural image generation, becoming a prominent research trend across various vertical domains. However, in the remote sensing field, there is still a lack of research on large-scale text-to-image (text2image) generation technology. Existing remote sensing image-text datasets are small in scale and confined to specific geographic areas and scene types. Besides, existing text2image methods have struggled to achieve global-scale, multi-resolution controllable, and unbounded image generation. To address these challenges, this paper presents two key contributions: the Git-10M dataset and the Text2Earth foundation model. Git-10M is a global-scale image-text dataset comprising 10 million image-text pairs, 5 times larger than the previous largest one. The dataset covers a wide range of geographic scenes and contains resolution information, significantly surpassing existing datasets in both size and diversity. Building on Git-10M, we propose Text2Earth, a 1.3 billion parameter generative foundation model based on the diffusion framework to model global-scale remote sensing scenes. Text2Earth integrates a resolution guidance mechanism, enabling users to specify image resolutions. A dynamic condition adaptation strategy is proposed for training and inference to improve image quality. Text2Earth excels in zero-shot text2image generation and demonstrates robust generalization and flexibility across multiple tasks, including unbounded scene construction, image editing, and cross-modal image generation. This robust capability surpasses previous models restricted to the basic fixed size and limited scene types. On the previous benchmark dataset, Text2Earth outperforms previous models with an improvement of +26.23 FID and +20.95% Zero-shot Cls-OA metric.Our project page is \url{https://chen-yang-liu.github.io/Text2Earth}
Authors: Jiaxin Ge, Zora Zhiruo Wang, Xuhui Zhou, Yi-Hao Peng, Sanjay Subramanian, Qinyue Tan, Maarten Sap, Alane Suhr, Daniel Fried, Graham Neubig, Trevor Darrell
Abstract: Designing structured visuals such as presentation slides is essential for communicative needs, necessitating both content creation and visual planning skills. In this work, we tackle the challenge of automated slide generation, where models produce slide presentations from natural language (NL) instructions. We first introduce the SlidesBench benchmark, the first benchmark for slide generation with 7k training and 585 testing examples derived from 310 slide decks across 10 domains. SlidesBench supports evaluations that are (i)reference-based to measure similarity to a target slide, and (ii)reference-free to measure the design quality of generated slides alone. We benchmark end-to-end image generation and program generation methods with a variety of models, and find that programmatic methods produce higher-quality slides in user-interactable formats. Built on the success of program generation, we create AutoPresent, an 8B Llama-based model trained on 7k pairs of instructions paired with code for slide generation, and achieve results comparable to the closed-source model GPT-4o. We further explore iterative design refinement where the model is tasked to self-refine its own output, and we found that this process improves the slide's quality. We hope that our work will provide a basis for future work on generating structured visuals.
Authors: Emily Johnson, Noah Wilson
Abstract: Text-to-image generation has witnessed significant advancements with the integration of Large Vision-Language Models (LVLMs), yet challenges remain in aligning complex textual descriptions with high-quality, visually coherent images. This paper introduces the Vision-Language Aligned Diffusion (VLAD) model, a generative framework that addresses these challenges through a dual-stream strategy combining semantic alignment and hierarchical diffusion. VLAD utilizes a Contextual Composition Module (CCM) to decompose textual prompts into global and local representations, ensuring precise alignment with visual features. Furthermore, it incorporates a multi-stage diffusion process with hierarchical guidance to generate high-fidelity images. Experiments conducted on MARIO-Eval and INNOVATOR-Eval benchmarks demonstrate that VLAD significantly outperforms state-of-the-art methods in terms of image quality, semantic alignment, and text rendering accuracy. Human evaluations further validate the superior performance of VLAD, making it a promising approach for text-to-image generation in complex scenarios.
Authors: Mallika Garg, Debashis Ghosh, Pyari Mohan Pradhan
Abstract: Dynamic gesture recognition is one of the challenging research areas due to variations in pose, size, and shape of the signer's hand. In this letter, Multiscaled Multi-Head Attention Video Transformer Network (MsMHA-VTN) for dynamic hand gesture recognition is proposed. A pyramidal hierarchy of multiscale features is extracted using the transformer multiscaled head attention model. The proposed model employs different attention dimensions for each head of the transformer which enables it to provide attention at the multiscale level. Further, in addition to single modality, recognition performance using multiple modalities is examined. Extensive experiments demonstrate the superior performance of the proposed MsMHA-VTN with an overall accuracy of 88.22\% and 99.10\% on NVGesture and Briareo datasets, respectively.
Authors: Hao Wang, Xiwen Chen, Ashish Bastola, Jiayou Qin, Abolfazl Razi
Abstract: The emergence of generative AI and controllable diffusion has made image-to-image synthesis increasingly practical and efficient. However, when input images exhibit low entropy and sparse, the inherent characteristics of diffusion models often result in limited diversity. This constraint significantly interferes with data augmentation. To address this, we propose Diffusion Prism, a training-free framework that efficiently transforms binary masks into realistic and diverse samples while preserving morphological features. We explored that a small amount of artificial noise will significantly assist the image-denoising process. To prove this novel mask-to-image concept, we use nano-dendritic patterns as an example to demonstrate the merit of our method compared to existing controllable diffusion models. Furthermore, we extend the proposed framework to other biological patterns, highlighting its potential applications across various fields.
Authors: Omid Saghatchian, Atiyeh Gh. Moghadam, Ahmad Nickabadi
Abstract: Diffusion models have emerged as a promising approach for generating high-quality, high-dimensional images. Nevertheless, these models are hindered by their high computational cost and slow inference, partly due to the quadratic computational complexity of the self-attention mechanisms with respect to input size. Various approaches have been proposed to address this drawback. One such approach focuses on reducing the number of tokens fed into the self-attention, known as token merging (ToMe). In our method, which is called cached adaptive token merging(CA-ToMe), we calculate the similarity between tokens and then merge the r proportion of the most similar tokens. However, due to the repetitive patterns observed in adjacent steps and the variation in the frequency of similarities, we aim to enhance this approach by implementing an adaptive threshold for merging tokens and adding a caching mechanism that stores similar pairs across several adjacent steps. Empirical results demonstrate that our method operates as a training-free acceleration method, achieving a speedup factor of 1.24 in the denoising process while maintaining the same FID scores compared to existing approaches.
Authors: Wenqi Zhang, Hang Zhang, Xin Li, Jiashuo Sun, Yongliang Shen, Weiming Lu, Deli Zhao, Yueting Zhuang, Lidong Bing
Abstract: Compared to image-text pair data, interleaved corpora enable Vision-Language Models (VLMs) to understand the world more naturally like humans. However, such existing datasets are crawled from webpage, facing challenges like low knowledge density, loose image-text relations, and poor logical coherence between images. On the other hand, the internet hosts vast instructional videos (e.g., online geometry courses) that are widely used by humans to learn foundational subjects, yet these valuable resources remain underexplored in VLM training. In this paper, we introduce a high-quality \textbf{multimodal textbook} corpus with richer foundational knowledge for VLM pretraining. It collects over 2.5 years of instructional videos, totaling 22,000 class hours. We first use an LLM-proposed taxonomy to systematically gather instructional videos. Then we progressively extract and refine visual (keyframes), audio (ASR), and textual knowledge (OCR) from the videos, and organize as an image-text interleaved corpus based on temporal order. Compared to its counterparts, our video-centric textbook offers more coherent context, richer knowledge, and better image-text alignment. Experiments demonstrate its superb pretraining performance, particularly in knowledge- and reasoning-intensive tasks like ScienceQA and MathVista. Moreover, VLMs pre-trained on our textbook exhibit outstanding interleaved context awareness, leveraging visual and textual cues in their few-shot context for task solving~\footnote{Our code are available at \url{https://github.com/DAMO-NLP-SG/multimodal_textbook}}.
Authors: Sepehr Dehdashtian, Gautam Sreekumar, Vishnu Naresh Boddeti
Abstract: Images generated by text-to-image (T2I) models often exhibit visual biases and stereotypes of concepts such as culture and profession. Existing quantitative measures of stereotypes are based on statistical parity that does not align with the sociological definition of stereotypes and, therefore, incorrectly categorizes biases as stereotypes. Instead of oversimplifying stereotypes as biases, we propose a quantitative measure of stereotypes that aligns with its sociological definition. We then propose OASIS to measure the stereotypes in a generated dataset and understand their origins within the T2I model. OASIS includes two scores to measure stereotypes from a generated image dataset: (M1) Stereotype Score to measure the distributional violation of stereotypical attributes, and (M2) WALS to measure spectral variance in the images along a stereotypical attribute. OASIS also includes two methods to understand the origins of stereotypes in T2I models: (U1) StOP to discover attributes that the T2I model internally associates with a given concept, and (U2) SPI to quantify the emergence of stereotypical attributes in the latent space of the T2I model during image generation. Despite the considerable progress in image fidelity, using OASIS, we conclude that newer T2I models such as FLUX.1 and SDv3 contain strong stereotypical predispositions about concepts and still generate images with widespread stereotypical attributes. Additionally, the quantity of stereotypes worsens for nationalities with lower Internet footprints.
Authors: Daniel Silver, Ron Kimmel
Abstract: In the field of video compression, the pursuit for better quality at lower bit rates remains a long-lasting goal. Recent developments have demonstrated the potential of Implicit Neural Representation (INR) as a promising alternative to traditional transform-based methodologies. Video INRs can be roughly divided into frame-wise and pixel-wise methods according to the structure the network outputs. While the pixel-based methods are better for upsampling and parallelization, frame-wise methods demonstrated better performance. We introduce CoordFlow, a novel pixel-wise INR for video compression. It yields state-of-the-art results compared to other pixel-wise INRs and on-par performance compared to leading frame-wise techniques. The method is based on the separation of the visual information into visually consistent layers, each represented by a dedicated network that compensates for the layer's motion. When integrated, a byproduct is an unsupervised segmentation of video sequence. Objects motion trajectories are implicitly utilized to compensate for visual-temporal redundancies. Additionally, the proposed method provides inherent video upsampling, stabilization, inpainting, and denoising capabilities.
Authors: Ao Gao, Luosong Guo, Tao Chen, Zhao Wang, Ying Tai, Jian Yang, Zhenyu Zhang
Abstract: 3D Gaussian Splatting (3DGS) techniques have achieved satisfactory 3D scene representation. Despite their impressive performance, they confront challenges due to the limitation of structure-from-motion (SfM) methods on acquiring accurate scene initialization, or the inefficiency of densification strategy. In this paper, we introduce a novel framework EasySplat to achieve high-quality 3DGS modeling. Instead of using SfM for scene initialization, we employ a novel method to release the power of large-scale pointmap approaches. Specifically, we propose an efficient grouping strategy based on view similarity, and use robust pointmap priors to obtain high-quality point clouds and camera poses for 3D scene initialization. After obtaining a reliable scene structure, we propose a novel densification approach that adaptively splits Gaussian primitives based on the average shape of neighboring Gaussian ellipsoids, utilizing KNN scheme. In this way, the proposed method tackles the limitation on initialization and optimization, leading to an efficient and accurate 3DGS modeling. Extensive experiments demonstrate that EasySplat outperforms the current state-of-the-art (SOTA) in handling novel view synthesis.
Authors: Zhaoyu Chen, Haijing Guo, Kaixun Jiang, Jiyuan Fu, Xinyu Zhou, Dingkang Yang, Hao Tang, Bo Li, Wenqiang Zhang
Abstract: Deep neural networks are vulnerable to adversarial examples that exhibit transferability across various models. Numerous approaches are proposed to enhance the transferability of adversarial examples, including advanced optimization, data augmentation, and model modifications. However, these methods still show limited transferability, particularly in cross-architecture scenarios, such as from CNN to ViT. To achieve high transferability, we propose a technique termed Spatial Adversarial Alignment (SAA), which employs an alignment loss and leverages a witness model to fine-tune the surrogate model. Specifically, SAA consists of two key parts: spatial-aware alignment and adversarial-aware alignment. First, we minimize the divergences of features between the two models in both global and local regions, facilitating spatial alignment. Second, we introduce a self-adversarial strategy that leverages adversarial examples to impose further constraints, aligning features from an adversarial perspective. Through this alignment, the surrogate model is trained to concentrate on the common features extracted by the witness model. This facilitates adversarial attacks on these shared features, thereby yielding perturbations that exhibit enhanced transferability. Extensive experiments on various architectures on ImageNet show that aligned surrogate models based on SAA can provide higher transferable adversarial examples, especially in cross-architecture attacks.
Authors: Anna Grim, Jayaram Chandrashekar, Uygar Sumbul
Abstract: Reconstructing the intricate local morphology of neurons and their long-range projecting axons can address many connectivity related questions in neuroscience. The main bottleneck in connectomics pipelines is correcting topological errors, as multiple entangled neuronal arbors is a challenging instance segmentation problem. More broadly, segmentation of curvilinear, filamentous structures continues to pose significant challenges. To address this problem, we extend the notion of simple points from digital topology to connected sets of voxels (i.e. supervoxels) and propose a topology-aware neural network segmentation method with minimal computational overhead. We demonstrate its effectiveness on a new public dataset of 3-d light microscopy images of mouse brains, along with the benchmark datasets DRIVE, ISBI12, and CrackTree.
Authors: Ziyang Chen, Yongjun Zhang, Wenting Li, Bingshu Wang, Yabo Wu, Yong Zhao, C. L. Philip Chen
Abstract: In light of the advancements in transformer technology, extant research posits the construction of stereo transformers as a potential solution to the binocular stereo matching challenge. However, constrained by the low-rank bottleneck and quadratic complexity of attention mechanisms, stereo transformers still fail to demonstrate sufficient nonlinear expressiveness within a reasonable inference time. The lack of focus on key homonymous points renders the representations of such methods vulnerable to challenging conditions, including reflections and weak textures. Furthermore, a slow computing speed is not conducive to the application. To overcome these difficulties, we present the \textbf{H}adamard \textbf{A}ttention \textbf{R}ecurrent Stereo \textbf{T}ransformer (HART) that incorporates the following components: 1) For faster inference, we present a Hadamard product paradigm for the attention mechanism, achieving linear computational complexity. 2) We designed a Dense Attention Kernel (DAK) to amplify the differences between relevant and irrelevant feature responses. This allows HART to focus on important details. DAK also converts zero elements to non-zero elements to mitigate the reduced expressiveness caused by the low-rank bottleneck. 3) To compensate for the spatial and channel interaction missing in the Hadamard product, we propose MKOI to capture both global and local information through the interleaving of large and small kernel convolutions. Experimental results demonstrate the effectiveness of our HART. In reflective area, HART ranked \textbf{1st} on the KITTI 2012 benchmark among all published methods at the time of submission. Code is available at \url{https://github.com/ZYangChen/HART}.
Authors: Huashan Chen, Yifan Xu, Yue Feng, Ming Jian, Feng Liu, Pengfei Hu, Kebin Peng, Sen He, Zi Wang
Abstract: Biometrics authentication has become increasingly popular due to its security and convenience; however, traditional biometrics are becoming less desirable in scenarios such as new mobile devices, Virtual Reality, and Smart Vehicles. For example, while face authentication is widely used, it suffers from significant privacy concerns. The collection of complete facial data makes it less desirable for privacy-sensitive applications. Lip authentication, on the other hand, has emerged as a promising biometrics method. However, existing lip-based authentication methods heavily depend on static lip shape when the mouth is closed, which can be less robust due to lip shape dynamic motion and can barely work when the user is speaking. In this paper, we revisit the nature of lip biometrics and extract shape-independent features from the lips. We study the dynamic characteristics of lip biometrics based on articulator motion. Building on the knowledge, we propose a system for shape-independent continuous authentication via lip articulator dynamics. This system enables robust, shape-independent and continuous authentication, making it particularly suitable for scenarios with high security and privacy requirements. We conducted comprehensive experiments in different environments and attack scenarios and collected a dataset of 50 subjects. The results indicate that our system achieves an overall accuracy of 99.06% and demonstrates robustness under advanced mimic attacks and AI deepfake attacks, making it a viable solution for continuous biometric authentication in various applications.
Authors: Jingkai Sun, Qiang Zhang, Jiaxu Wang, Jiahang Cao, Renjing Xu
Abstract: Dynamic vision sensors (DVS) are bio-inspired devices that capture visual information in the form of asynchronous events, which encode changes in pixel intensity with high temporal resolution and low latency. These events provide rich motion cues that can be exploited for various computer vision tasks, such as action recognition. However, most existing DVS-based action recognition methods lose temporal information during data transformation or suffer from noise and outliers caused by sensor imperfections or environmental factors. To address these challenges, we propose a novel framework that preserves and exploits the spatiotemporal structure of event data for action recognition. Our framework consists of two main components: 1) a point-wise event masked autoencoder (MAE) that learns a compact and discriminative representation of event patches by reconstructing them from masked raw event camera points data; 2) an improved event points patch generation algorithm that leverages an event data inlier model and point-wise data augmentation techniques to enhance the quality and diversity of event points patches. To the best of our knowledge, our approach introduces the pre-train method into event camera raw points data for the first time, and we propose a novel event points patch embedding to utilize transformer-based models on event cameras.
Authors: Linhao Huang, Xue Jiang, Zhiqiang Wang, Wentao Mo, Xi Xiao, Bo Han, Yongjie Yin, Feng Zheng
Abstract: Video-based multimodal large language models (V-MLLMs) have shown vulnerability to adversarial examples in video-text multimodal tasks. However, the transferability of adversarial videos to unseen models--a common and practical real world scenario--remains unexplored. In this paper, we pioneer an investigation into the transferability of adversarial video samples across V-MLLMs. We find that existing adversarial attack methods face significant limitations when applied in black-box settings for V-MLLMs, which we attribute to the following shortcomings: (1) lacking generalization in perturbing video features, (2) focusing only on sparse key-frames, and (3) failing to integrate multimodal information. To address these limitations and deepen the understanding of V-MLLM vulnerabilities in black-box scenarios, we introduce the Image-to-Video MLLM (I2V-MLLM) attack. In I2V-MLLM, we utilize an image-based multimodal model (IMM) as a surrogate model to craft adversarial video samples. Multimodal interactions and temporal information are integrated to disrupt video representations within the latent space, improving adversarial transferability. In addition, a perturbation propagation technique is introduced to handle different unknown frame sampling strategies. Experimental results demonstrate that our method can generate adversarial examples that exhibit strong transferability across different V-MLLMs on multiple video-text multimodal tasks. Compared to white-box attacks on these models, our black-box attacks (using BLIP-2 as surrogate model) achieve competitive performance, with average attack success rates of 55.48% on MSVD-QA and 58.26% on MSRVTT-QA for VideoQA tasks, respectively. Our code will be released upon acceptance.
Authors: Tao Feng, Wei Li, DiDi Zhu, Hangjie Yuan, Wendi Zheng, Dan Zhang, Jie Tang
Abstract: Backpropagation provides a generalized configuration for overcoming catastrophic forgetting. Like, SGD and Adam are commonly used for weight updates in continual learning and continual pre-training. In practice, permission to access gradient information is not always granted (the gradient ban), such as black-box APIs, hardware limitations, and non-differentiable systems. To bridge this gap, we introduce the first benchmark ZeroFlow to evaluate gradient-free optimization algorithms for overcoming forgetting. This benchmark examines a suite of forward pass methods across multiple methods, forgetting scenarios, and datasets. We find that forward passes alone are enough to overcome forgetting. Our findings reveal new optimization principles that highlight the potential of forward-pass in mitigating forgetting, managing task conflicts, and reducing memory demands, alongside novel enhancements that further mitigate forgetting with just one forward pass. This work provides essential insights and tools for advancing forward pass methods to overcome forgetting.
Authors: Song Zhang, Zhiwei Wei, Wenjia Xu, Lili Zhang, Yang Wang, Jinming Zhang, Junyi Liu
Abstract: 3D terrain reconstruction with remote sensing imagery achieves cost-effective and large-scale earth observation and is crucial for safeguarding natural disasters, monitoring ecological changes, and preserving the environment.Recently, learning-based multi-view stereo~(MVS) methods have shown promise in this task. However, these methods simply modify the general learning-based MVS framework for height estimation, which overlooks the terrain characteristics and results in insufficient accuracy. Considering that the Earth's surface generally undulates with no drastic changes and can be measured by slope, integrating slope considerations into MVS frameworks could enhance the accuracy of terrain reconstructions. To this end, we propose an end-to-end slope-aware height estimation network named TS-SatMVSNet for large-scale remote sensing terrain reconstruction.To effectively obtain the slope representation, drawing from mathematical gradient concepts, we innovatively proposed a height-based slope calculation strategy to first calculate a slope map from a height map to measure the terrain undulation. To fully integrate slope information into the MVS pipeline, we separately design two slope-guided modules to enhance reconstruction outcomes at both micro and macro levels. Specifically, at the micro level, we designed a slope-guided interval partition module for refined height estimation using slope values. At the macro level, a height correction module is proposed, using a learnable Gaussian smoothing operator to amend the inaccurate height values. Additionally, to enhance the efficacy of height estimation, we proposed a slope direction loss for implicitly optimizing height estimation results. Extensive experiments on the WHU-TLC dataset and MVS3D dataset show that our proposed method achieves state-of-the-art performance and demonstrates competitive generalization ability.
Authors: Jiang Shang, Yuanmeng Wu, Xiaoxiang Han, Xi Chen, Qi Zhang
Abstract: Accurate and robust ultrasound image segmentation is critical for computer-aided diagnostic systems. Nevertheless, the inherent challenges of ultrasound imaging, such as blurry boundaries and speckle noise, often cause traditional segmentation methods to struggle with performance. Despite recent advancements in universal image segmentation, such as the Segment Anything Model, existing interactive segmentation methods still suffer from inefficiency and lack of specialization. These methods rely heavily on extensive accurate manual or random sampling prompts for interaction, necessitating numerous prompts and iterations to reach satisfactory performance. In response to this challenge, we propose the Evidential Uncertainty-Guided Interactive Segmentation (EUGIS), an end-to-end, efficient tiered interactive segmentation paradigm based on evidential uncertainty estimation for ultrasound image segmentation. Specifically, EUGIS harnesses evidence-based uncertainty estimation, grounded in Dempster-Shafer theory and Subjective Logic, to gauge the level of uncertainty in the predictions of model for different regions. By prioritizing sampling the high-uncertainty region, our method can effectively simulate the interactive behavior of well-trained radiologists, enhancing the targeted of sampling while reducing the number of prompts and iterations required.Additionally, we propose a trainable calibration mechanism for uncertainty estimation, which can further optimize the boundary between certainty and uncertainty, thereby enhancing the confidence of uncertainty estimation.
Authors: Hong Zhang, Zhongjie Duan, Xingjun Wang, Yingda Chen, Yu Zhang
Abstract: Recent advancements in diffusion models have significantly advanced text-to-image generation, yet global text prompts alone remain insufficient for achieving fine-grained control over individual entities within an image. To address this limitation, we present EliGen, a novel framework for Entity-Level controlled Image Generation. We introduce regional attention, a mechanism for diffusion transformers that requires no additional parameters, seamlessly integrating entity prompts and arbitrary-shaped spatial masks. By contributing a high-quality dataset with fine-grained spatial and semantic entity-level annotations, we train EliGen to achieve robust and accurate entity-level manipulation, surpassing existing methods in both positional control precision and image quality. Additionally, we propose an inpainting fusion pipeline, extending EliGen to multi-entity image inpainting tasks. We further demonstrate its flexibility by integrating it with community models such as IP-Adapter and MLLM, unlocking new creative possibilities. The source code, dataset, and model will be released publicly.
Authors: Jiwei Shan, Zeyu Cai, Cheng-Tai Hsieh, Shing Shin Cheng, Hesheng Wang
Abstract: Efficient and high-fidelity reconstruction of deformable surgical scenes is a critical yet challenging task. Building on recent advancements in 3D Gaussian splatting, current methods have seen significant improvements in both reconstruction quality and rendering speed. However, two major limitations remain: (1) difficulty in handling irreversible dynamic changes, such as tissue shearing, which are common in surgical scenes; and (2) the lack of hierarchical modeling for surgical scene deformation, which reduces rendering speed. To address these challenges, we introduce EH-SurGS, an efficient and high-fidelity reconstruction algorithm for deformable surgical scenes. We propose a deformation modeling approach that incorporates the life cycle of 3D Gaussians, effectively capturing both regular and irreversible deformations, thus enhancing reconstruction quality. Additionally, we present an adaptive motion hierarchy strategy that distinguishes between static and deformable regions within the surgical scene. This strategy reduces the number of 3D Gaussians passing through the deformation field, thereby improving rendering speed. Extensive experiments demonstrate that our method surpasses existing state-of-the-art approaches in both reconstruction quality and rendering speed. Ablation studies further validate the effectiveness and necessity of our proposed components. We will open-source our code upon acceptance of the paper.
Authors: Teng Li, Xingjun Ma, Yu-Gang Jiang
Abstract: Transferable adversarial examples highlight the vulnerability of deep neural networks (DNNs) to imperceptible perturbations across various real-world applications. While there have been notable advancements in untargeted transferable attacks, targeted transferable attacks remain a significant challenge. In this work, we focus on generative approaches for targeted transferable attacks. Current generative attacks focus on reducing overfitting to surrogate models and the source data domain, but they often overlook the importance of enhancing transferability through additional semantics. To address this issue, we introduce a novel plug-and-play module into the general generator architecture to enhance adversarial transferability. Specifically, we propose a \emph{Semantic Injection Module} (SIM) that utilizes the semantics contained in an additional guiding image to improve transferability. The guiding image provides a simple yet effective method to incorporate target semantics from the target class to create targeted and highly transferable attacks. Additionally, we propose new loss formulations that can integrate the semantic injection module more effectively for both targeted and untargeted attacks. We conduct comprehensive experiments under both targeted and untargeted attack settings to demonstrate the efficacy of our proposed approach.
Authors: Xiusheng Xu, Lei Qi, Jingyang Zhou, Xin Geng
Abstract: Source-Free Domain Generalization (SFDG) aims to develop a model that performs on unseen domains without relying on any source domains. However, the implementation remains constrained due to the unavailability of training data. Research on SFDG focus on knowledge transfer of multi-modal models and style synthesis based on joint space of multiple modalities, thus eliminating the dependency on source domain images. However, existing works primarily work for multi-domain and less-category configuration, but performance on multi-domain and multi-category configuration is relatively poor. In addition, the efficiency of style synthesis also deteriorates in multi-category scenarios. How to efficiently synthesize sufficiently diverse data and apply it to multi-category configuration is a direction with greater practical value. In this paper, we propose a method called BatStyler, which is utilized to improve the capability of style synthesis in multi-category scenarios. BatStyler consists of two modules: Coarse Semantic Generation and Uniform Style Generation modules. The Coarse Semantic Generation module extracts coarse-grained semantics to prevent the compression of space for style diversity learning in multi-category configuration, while the Uniform Style Generation module provides a template of styles that are uniformly distributed in space and implements parallel training. Extensive experiments demonstrate that our method exhibits comparable performance on less-category datasets, while surpassing state-of-the-art methods on multi-category datasets.
Authors: Dong Zhang, Kwang-Ting Cheng
Abstract: Thanks to the recent achievements in task-driven image quality enhancement (IQE) models like ESTR, the image enhancement model and the visual recognition model can mutually enhance each other's quantitation while producing high-quality processed images that are perceivable by our human vision systems. However, existing task-driven IQE models tend to overlook an underlying fact -- different levels of vision tasks have varying and sometimes conflicting requirements of image features. To address this problem, this paper proposes a generalized gradient promotion (GradProm) training strategy for task-driven IQE of medical images. Specifically, we partition a task-driven IQE system into two sub-models, i.e., a mainstream model for image enhancement and an auxiliary model for visual recognition. During training, GradProm updates only parameters of the image enhancement model using gradients of the visual recognition model and the image enhancement model, but only when gradients of these two sub-models are aligned in the same direction, which is measured by their cosine similarity. In case gradients of these two sub-models are not in the same direction, GradProm only uses the gradient of the image enhancement model to update its parameters. Theoretically, we have proved that the optimization direction of the image enhancement model will not be biased by the auxiliary visual recognition model under the implementation of GradProm. Empirically, extensive experimental results on four public yet challenging medical image datasets demonstrated the superior performance of GradProm over existing state-of-the-art methods.
Authors: Zitong Xu, Huiyu Duan, Guangji Ma, Liu Yang, Jiarui Wang, Qingbo Wu, Xiongkuo Min, Guangtao Zhai, Patrick Le Callet
Abstract: Image composition involves extracting a foreground object from one image and pasting it into another image through Image harmonization algorithms (IHAs), which aim to adjust the appearance of the foreground object to better match the background. Existing image quality assessment (IQA) methods may fail to align with human visual preference on image harmonization due to the insensitivity to minor color or light inconsistency. To address the issue and facilitate the advancement of IHAs, we introduce the first Image Quality Assessment Database for image Harmony evaluation (HarmonyIQAD), which consists of 1,350 harmonized images generated by 9 different IHAs, and the corresponding human visual preference scores. Based on this database, we propose a Harmony Image Quality Assessment (HarmonyIQA), to predict human visual preference for harmonized images. Extensive experiments show that HarmonyIQA achieves state-of-the-art performance on human visual preference evaluation for harmonized images, and also achieves competing results on traditional IQA tasks. Furthermore, cross-dataset evaluation also shows that HarmonyIQA exhibits better generalization ability than self-supervised learning-based IQA methods. Both HarmonyIQAD and HarmonyIQA will be made publicly available upon paper publication.
Authors: Xuan Yu, Yuxuan Xie, Yili Liu, Haojian Lu, Rong Xiong, Yiyi Liao, Yue Wang
Abstract: Open-vocabulary panoptic reconstruction offers comprehensive scene understanding, enabling advances in embodied robotics and photorealistic simulation. In this paper, we propose PanopticRecon++, an end-to-end method that formulates panoptic reconstruction through a novel cross-attention perspective. This perspective models the relationship between 3D instances (as queries) and the scene's 3D embedding field (as keys) through their attention map. Unlike existing methods that separate the optimization of queries and keys or overlook spatial proximity, PanopticRecon++ introduces learnable 3D Gaussians as instance queries. This formulation injects 3D spatial priors to preserve proximity while maintaining end-to-end optimizability. Moreover, this query formulation facilitates the alignment of 2D open-vocabulary instance IDs across frames by leveraging optimal linear assignment with instance masks rendered from the queries. Additionally, we ensure semantic-instance segmentation consistency by fusing query-based instance segmentation probabilities with semantic probabilities in a novel panoptic head supervised by a panoptic loss. During training, the number of instance query tokens dynamically adapts to match the number of objects. PanopticRecon++ shows competitive performance in terms of 3D and 2D segmentation and reconstruction performance on both simulation and real-world datasets, and demonstrates a user case as a robot simulator. Our project website is at: https://yuxuan1206.github.io/panopticrecon_pp/
Authors: Jian Lang, Zhangtao Cheng, Ting Zhong, Fan Zhou
Abstract: Multimodal learning with incomplete modality is practical and challenging. Recently, researchers have focused on enhancing the robustness of pre-trained MultiModal Transformers (MMTs) under missing modality conditions by applying learnable prompts. However, these prompt-based methods face several limitations: (1) incomplete modalities provide restricted modal cues for task-specific inference, (2) dummy imputation for missing content causes information loss and introduces noise, and (3) static prompts are instance-agnostic, offering limited knowledge for instances with various missing conditions. To address these issues, we propose RAGPT, a novel Retrieval-AuGmented dynamic Prompt Tuning framework. RAGPT comprises three modules: (I) the multi-channel retriever, which identifies similar instances through a within-modality retrieval strategy, (II) the missing modality generator, which recovers missing information using retrieved contexts, and (III) the context-aware prompter, which captures contextual knowledge from relevant instances and generates dynamic prompts to largely enhance the MMT's robustness. Extensive experiments conducted on three real-world datasets show that RAGPT consistently outperforms all competitive baselines in handling incomplete modality problems. The code of our work and prompt-based baselines is available at https://github.com/Jian-Lang/RAGPT.
Authors: Zhenyu Li, Wenqing Cui, Shariq Farooq Bhat, Peter Wonka
Abstract: While current high-resolution depth estimation methods achieve strong results, they often suffer from computational inefficiencies due to reliance on heavyweight models and multiple inference steps, increasing inference time. To address this, we introduce PatchRefiner V2 (PRV2), which replaces heavy refiner models with lightweight encoders. This reduces model size and inference time but introduces noisy features. To overcome this, we propose a Coarse-to-Fine (C2F) module with a Guided Denoising Unit for refining and denoising the refiner features and a Noisy Pretraining strategy to pretrain the refiner branch to fully exploit the potential of the lightweight refiner branch. Additionally, we introduce a Scale-and-Shift Invariant Gradient Matching (SSIGM) loss to enhance synthetic-to-real domain transfer. PRV2 outperforms state-of-the-art depth estimation methods on UnrealStereo4K in both accuracy and speed, using fewer parameters and faster inference. It also shows improved depth boundary delineation on real-world datasets like CityScape, ScanNet++, and KITTI, demonstrating its versatility across domains.
Authors: Feng Han, Kai Chen, Chao Gong, Zhipeng Wei, Jingjing Chen, Yu-Gang Jiang
Abstract: The exceptional generative capability of text-to-image models has raised substantial safety concerns regarding the generation of Not-Safe-For-Work (NSFW) content and potential copyright infringement. To address these concerns, previous methods safeguard the models by eliminating inappropriate concepts. Nonetheless, these models alter the parameters of the backbone network and exert considerable influences on the structural (low-frequency) components of the image, which undermines the model's ability to retain non-target concepts. In this work, we propose our Dual encoder Modulation network (DuMo), which achieves precise erasure of inappropriate target concepts with minimum impairment to non-target concepts. In contrast to previous methods, DuMo employs the Eraser with PRior Knowledge (EPR) module which modifies the skip connection features of the U-NET and primarily achieves concept erasure on details (high-frequency) components of the image. To minimize the damage to non-target concepts during erasure, the parameters of the backbone U-NET are frozen and the prior knowledge from the original skip connection features is introduced to the erasure process. Meanwhile, the phenomenon is observed that distinct erasing preferences for the image structure and details are demonstrated by the EPR at different timesteps and layers. Therefore, we adopt a novel Time-Layer MOdulation process (TLMO) that adjusts the erasure scale of EPR module's outputs across different layers and timesteps, automatically balancing the erasure effects and model's generative ability. Our method achieves state-of-the-art performance on Explicit Content Erasure, Cartoon Concept Removal and Artistic Style Erasure, clearly outperforming alternative methods. Code is available at https://github.com/Maplebb/DuMo
Authors: Xinyang Huang, Chuang Zhu, Ruiying Ren, Shengjie Liu, Tiejun Huang
Abstract: Semi-supervised domain adaptation (SSDA) has been extensively researched due to its ability to improve classification performance and generalization ability of models by using a small amount of labeled data on the target domain. However, existing methods cannot effectively adapt to the target domain due to difficulty in fully learning rich and complex target semantic information and relationships. In this paper, we propose a novel SSDA learning framework called semantic regularization learning (SERL), which captures the target semantic information from multiple perspectives of regularization learning to achieve adaptive fine-tuning of the source pre-trained model on the target domain. SERL includes three robust semantic regularization techniques. Firstly, semantic probability contrastive regularization (SPCR) helps the model learn more discriminative feature representations from a probabilistic perspective, using semantic information on the target domain to understand the similarities and differences between samples. Additionally, adaptive weights in SPCR can help the model learn the semantic distribution correctly through the probabilities of different samples. To further comprehensively understand the target semantic distribution, we introduce hard-sample mixup regularization (HMR), which uses easy samples as guidance to mine the latent target knowledge contained in hard samples, thereby learning more complete and complex target semantic knowledge. Finally, target prediction regularization (TPR) regularizes the target predictions of the model by maximizing the correlation between the current prediction and the past learned objective, thereby mitigating the misleading of semantic information caused by erroneous pseudo-labels. Extensive experiments on three benchmark datasets demonstrate that our SERL method achieves state-of-the-art performance.
Authors: Sihan Wang, Shangqi Gao, Fuping Wu, Xiahai Zhuang
Abstract: Image decomposition aims to analyze an image into elementary components, which is essential for numerous downstream tasks and also by nature provides certain interpretability to the analysis. Deep learning can be powerful for such tasks, but surprisingly their combination with a focus on interpretability and generalizability is rarely explored. In this work, we introduce a novel framework for interpretable deep image decomposition, combining hierarchical Bayesian modeling and deep learning to create an architecture-modularized and model-generalizable deep neural network (DNN). The proposed framework includes three steps: (1) hierarchical Bayesian modeling of image decomposition, (2) transforming the inference problem into optimization tasks, and (3) deep inference via a modularized Bayesian DNN. We further establish a theoretical connection between the loss function and the generalization error bound, which inspires a new test-time adaptation approach for out-of-distribution scenarios. We instantiated the application using two downstream tasks, \textit{i.e.}, image denoising, and unsupervised anomaly detection, and the results demonstrated improved generalizability as well as interpretability of our methods. The source code will be released upon the acceptance of this paper.
Authors: Yang Yuxiang, Zeng Xinyi, Zeng Pinxian, Zu Chen, Yan Binyu, Zhou Jiliu, Wang Yan
Abstract: Multi-source Domain Adaptation (MDA) aims to transfer knowledge from multiple labeled source domains to an unlabeled target domain. Nevertheless, traditional methods primarily focus on achieving inter-domain alignment through sample-level constraints, such as Maximum Mean Discrepancy (MMD), neglecting three pivotal aspects: 1) the potential of data augmentation, 2) the significance of intra-domain alignment, and 3) the design of cluster-level constraints. In this paper, we introduce a novel hardness-driven strategy for MDA tasks, named "A3MDA" , which collectively considers these three aspects through Adaptive hardness quantification and utilization in both data Augmentation and domain Alignment.To achieve this, "A3MDA" progressively proposes three Adaptive Hardness Measurements (AHM), i.e., Basic, Smooth, and Comparative AHMs, each incorporating distinct mechanisms for diverse scenarios. Specifically, Basic AHM aims to gauge the instantaneous hardness for each source/target sample. Then, hardness values measured by Smooth AHM will adaptively adjust the intensity level of strong data augmentation to maintain compatibility with the model's generalization capacity.In contrast, Comparative AHM is designed to facilitate cluster-level constraints. By leveraging hardness values as sample-specific weights, the traditional MMD is enhanced into a weighted-clustered variant, strengthening the robustness and precision of inter-domain alignment. As for the often-neglected intra-domain alignment, we adaptively construct a pseudo-contrastive matrix by selecting harder samples based on the hardness rankings, enhancing the quality of pseudo-labels, and shaping a well-clustered target feature space. Experiments on multiple MDA benchmarks show that " A3MDA " outperforms other methods.
Authors: Vriksha Srihari, R. Bhavya, Shruti Jayaraman, V. Mary Anita Rajam
Abstract: While generative models such as text-to-image, large language models and text-to-video have seen significant progress, the extension to text-to-virtual-reality remains largely unexplored, due to a deficit in training data and the complexity of achieving realistic depth and motion in virtual environments. This paper proposes an approach to coalesce existing generative systems to form a stereoscopic virtual reality video from text. Carried out in three main stages, we start with a base text-to-image model that captures context from an input text. We then employ Stable Diffusion on the rudimentary image produced, to generate frames with enhanced realism and overall quality. These frames are processed with depth estimation algorithms to create left-eye and right-eye views, which are stitched side-by-side to create an immersive viewing experience. Such systems would be highly beneficial in virtual reality production, since filming and scene building often require extensive hours of work and post-production effort. We utilize image evaluation techniques, specifically Fr\'echet Inception Distance and CLIP Score, to assess the visual quality of frames produced for the video. These quantitative measures establish the proficiency of the proposed method. Our work highlights the exciting possibilities of using natural language-driven graphics in fields like virtual reality simulations.
Authors: Jiajun Deng, Tianyu He, Li Jiang, Tianyu Wang, Feras Dayoub, Ian Reid
Abstract: Current 3D Large Multimodal Models (3D LMMs) have shown tremendous potential in 3D-vision-based dialogue and reasoning. However, how to further enhance 3D LMMs to achieve fine-grained scene understanding and facilitate flexible human-agent interaction remains a challenging problem. In this work, we introduce 3D-LLaVA, a simple yet highly powerful 3D LMM designed to act as an intelligent assistant in comprehending, reasoning, and interacting with the 3D world. Unlike existing top-performing methods that rely on complicated pipelines-such as offline multi-view feature extraction or additional task-specific heads-3D-LLaVA adopts a minimalist design with integrated architecture and only takes point clouds as input. At the core of 3D-LLaVA is a new Omni Superpoint Transformer (OST), which integrates three functionalities: (1) a visual feature selector that converts and selects visual tokens, (2) a visual prompt encoder that embeds interactive visual prompts into the visual token space, and (3) a referring mask decoder that produces 3D masks based on text description. This versatile OST is empowered by the hybrid pretraining to obtain perception priors and leveraged as the visual connector that bridges the 3D data to the LLM. After performing unified instruction tuning, our 3D-LLaVA reports impressive results on various benchmarks. The code and model will be released to promote future exploration.
Authors: Lixiong Qin, Ning Jiang, Yang Zhang, Yuhan Qiu, Dingheng Zeng, Jiani Hu, Weihong Deng
Abstract: Existing deepfake analysis methods are primarily based on discriminative models, which significantly limit their application scenarios. This paper aims to explore interactive deepfake analysis by performing instruction tuning on multi-modal large language models (MLLMs). This will face challenges such as the lack of datasets and benchmarks, and low training efficiency. To address these issues, we introduce (1) a GPT-assisted data construction process resulting in an instruction-following dataset called DFA-Instruct, (2) a benchmark named DFA-Bench, designed to comprehensively evaluate the capabilities of MLLMs in deepfake detection, deepfake classification, and artifact description, and (3) construct an interactive deepfake analysis system called DFA-GPT, as a strong baseline for the community, with the Low-Rank Adaptation (LoRA) module. The dataset and code will be made available at https://github.com/lxq1000/DFA-Instruct to facilitate further research.
Authors: Chengrui Gao, Ziyuan Yang, Wei Jia, Lu Leng, Bob Zhang, Andrew Beng Jin Teoh
Abstract: Palmprint recognition has emerged as a prominent biometric technology, widely applied in diverse scenarios. Traditional handcrafted methods for palmprint recognition often fall short in representation capability, as they heavily depend on researchers' prior knowledge. Deep learning (DL) has been introduced to address this limitation, leveraging its remarkable successes across various domains. While existing surveys focus narrowly on specific tasks within palmprint recognition-often grounded in traditional methodologies-there remains a significant gap in comprehensive research exploring DL-based approaches across all facets of palmprint recognition. This paper bridges that gap by thoroughly reviewing recent advancements in DL-powered palmprint recognition. The paper systematically examines progress across key tasks, including region-of-interest segmentation, feature extraction, and security/privacy-oriented challenges. Beyond highlighting these advancements, the paper identifies current challenges and uncovers promising opportunities for future research. By consolidating state-of-the-art progress, this review serves as a valuable resource for researchers, enabling them to stay abreast of cutting-edge technologies and drive innovation in palmprint recognition.
Authors: Soumyaratna Debnath, Harish Katti, Shashikant Verma, Shanmuganathan Raman
Abstract: While 2D pose estimation has advanced our ability to interpret body movements in animals and primates, it is limited by the lack of depth information, constraining its application range. 3D pose estimation provides a more comprehensive solution by incorporating spatial depth, yet creating extensive 3D pose datasets for animals is challenging due to their dynamic and unpredictable behaviours in natural settings. To address this, we propose a hybrid approach that utilizes rigged avatars and the pipeline to generate synthetic datasets to acquire the necessary 3D annotations for training. Our method introduces a simple attention-based MLP network for converting 2D poses to 3D, designed to be independent of the input image to ensure scalability for poses in natural environments. Additionally, we identify that existing anatomical keypoint detectors are insufficient for accurate pose retargeting onto arbitrary avatars. To overcome this, we present a lookup table based on a deep pose estimation method using a synthetic collection of diverse actions rigged avatars perform. Our experiments demonstrate the effectiveness and efficiency of this lookup table-based retargeting approach. Overall, we propose a comprehensive framework with systematically synthesized datasets for lifting poses from 2D to 3D and then utilize this to re-target motion from wild settings onto arbitrary avatars.
Authors: Dat Nguyen, Marcella Astrid, Anis Kacem, Enjie Ghorbel, Djamila Aouada
Abstract: Detecting deepfake videos is highly challenging due to the complex intertwined spatial and temporal artifacts in forged sequences. Most recent approaches rely on binary classifiers trained on both real and fake data. However, such methods may struggle to focus on important artifacts, which can hinder their generalization capability. Additionally, these models often lack interpretability, making it difficult to understand how predictions are made. To address these issues, we propose FakeSTormer, offering two key contributions. First, we introduce a multi-task learning framework with additional spatial and temporal branches that enable the model to focus on subtle spatio-temporal artifacts. These branches also provide interpretability by highlighting video regions that may contain artifacts. Second, we propose a video-level data synthesis algorithm that generates pseudo-fake videos with subtle artifacts, providing the model with high-quality samples and ground truth data for our spatial and temporal branches. Extensive experiments on several challenging benchmarks demonstrate the competitiveness of our approach compared to recent state-of-the-art methods. The code is available at https://github.com/10Ring/FakeSTormer.
Authors: Yulun Wu, Han Huang, Wenyuan Zhang, Chao Deng, Ge Gao, Ming Gu, Yu-Shen Liu
Abstract: In recent years, reconstructing indoor scene geometry from multi-view images has achieved encouraging accomplishments. Current methods incorporate monocular priors into neural implicit surface models to achieve high-quality reconstructions. However, these methods require hundreds of images for scene reconstruction. When only a limited number of views are available as input, the performance of monocular priors deteriorates due to scale ambiguity, leading to the collapse of the reconstructed scene geometry. In this paper, we propose a new method, named Sparis, for indoor surface reconstruction from sparse views. Specifically, we investigate the impact of monocular priors on sparse scene reconstruction, introducing a novel prior based on inter-image matching information. Our prior offers more accurate depth information while ensuring cross-view matching consistency. Additionally, we employ an angular filter strategy and an epipolar matching weight function, aiming to reduce errors due to view matching inaccuracies, thereby refining the inter-image prior for improved reconstruction accuracy. The experiments conducted on widely used benchmarks demonstrate superior performance in sparse-view scene reconstruction.
Authors: Kyoungkook Kang, Gyujin Sim, Geonung Kim, Donguk Kim, Seungho Nam, Sunghyun Cho
Abstract: Layers have become indispensable tools for professional artists, allowing them to build a hierarchical structure that enables independent control over individual visual elements. In this paper, we propose LayeringDiff, a novel pipeline for the synthesis of layered images, which begins by generating a composite image using an off-the-shelf image generative model, followed by disassembling the image into its constituent foreground and background layers. By extracting layers from a composite image, rather than generating them from scratch, LayeringDiff bypasses the need for large-scale training to develop generative capabilities for individual layers. Furthermore, by utilizing a pretrained off-the-shelf generative model, our method can produce diverse contents and object scales in synthesized layers. For effective layer decomposition, we adapt a large-scale pretrained generative prior to estimate foreground and background layers. We also propose high-frequency alignment modules to refine the fine-details of the estimated layers. Our comprehensive experiments demonstrate that our approach effectively synthesizes layered images and supports various practical applications.
Authors: Yitong Zhu, Tangyao Li, Yuyang Wang
Abstract: Cybersickness remains a significant barrier to the widespread adoption of immersive virtual reality (VR) experiences, as it can greatly disrupt user engagement and comfort. Research has shown that cybersickness can significantly be reflected in head and eye tracking data, along with other physiological data (e.g., TMP, EDA, and BMP). Despite the application of deep learning techniques such as CNNs and LSTMs, these models often struggle to capture the complex interactions between multiple data modalities and lack the capacity for real-time inference, limiting their practical application. Addressing this gap, we propose a lightweight model that leverages a transformer-based encoder with sparse self-attention to process bio-signal features and a PP-TSN network for video feature extraction. These features are then integrated via a cross-modal fusion module, creating a video-aware bio-signal representation that supports cybersickness prediction based on both visual and bio-signal inputs. Our model, trained with a lightweight framework, was validated on a public dataset containing eye and head tracking data, physiological data, and VR video, and demonstrated state-of-the-art performance in cybersickness prediction, achieving a high accuracy of 93.13\% using only VR video inputs. These findings suggest that our approach not only enables effective, real-time cybersickness prediction but also addresses the longstanding issue of modality interaction in VR environments. This advancement provides a foundation for future research on multimodal data integration in VR, potentially leading to more personalized, comfortable and widely accessible VR experiences.
Authors: A. V. Subramanyam, Amil Bhagat, Milind Jain
Abstract: Consistency models have emerged as a promising alternative to diffusion models, offering high-quality generative capabilities through single-step sample generation. However, their application to multi-domain image translation tasks, such as cross-modal translation and low-light image enhancement remains largely unexplored. In this paper, we introduce Conditional Consistency Models (CCMs) for multi-domain image translation by incorporating additional conditional inputs. We implement these modifications by introducing task-specific conditional inputs that guide the denoising process, ensuring that the generated outputs retain structural and contextual information from the corresponding input domain. We evaluate CCMs on 10 different datasets demonstrating their effectiveness in producing high-quality translated images across multiple domains. Code is available at https://github.com/amilbhagat/Conditional-Consistency-Models.
URLs: https://github.com/amilbhagat/Conditional-Consistency-Models.
Authors: Muhammet Balcilar, Bharath Bhushan Damodaran, Karam Naser, Franck Galpin, Pierre Hellier
Abstract: End-to-end image and video codecs are becoming increasingly competitive, compared to traditional compression techniques that have been developed through decades of manual engineering efforts. These trainable codecs have many advantages over traditional techniques, such as their straightforward adaptation to perceptual distortion metrics and high performance in specific fields thanks to their learning ability. However, current state-of-the-art neural codecs do not fully exploit the benefits of vector quantization and the existence of the entropy gradient in decoding devices. In this paper, we propose to leverage these two properties (vector quantization and entropy gradient) to improve the performance of off-the-shelf codecs. Firstly, we demonstrate that using non-uniform scalar quantization cannot improve performance over uniform quantization. We thus suggest using predefined optimal uniform vector quantization to improve performance. Secondly, we show that the entropy gradient, available at the decoder, is correlated with the reconstruction error gradient, which is not available at the decoder. We therefore use the former as a proxy to enhance compression performance. Our experimental results show that these approaches save between 1 to 3% of the rate for the same quality across various pretrained methods. In addition, the entropy gradient based solution improves traditional codec performance significantly as well.
Authors: Zhiyao Wang, Xu Chen, Chengming Xu, Junwei Zhu, Xiaobin Hu, Jiangning Zhang, Chengjie Wang, Yuqi Liu, Yiyi Zhou, Rongrong Ji
Abstract: Face Restoration (FR) is a crucial area within image and video processing, focusing on reconstructing high-quality portraits from degraded inputs. Despite advancements in image FR, video FR remains relatively under-explored, primarily due to challenges related to temporal consistency, motion artifacts, and the limited availability of high-quality video data. Moreover, traditional face restoration typically prioritizes enhancing resolution and may not give as much consideration to related tasks such as facial colorization and inpainting. In this paper, we propose a novel approach for the Generalized Video Face Restoration (GVFR) task, which integrates video BFR, inpainting, and colorization tasks that we empirically show to benefit each other. We present a unified framework, termed as stable video face restoration (SVFR), which leverages the generative and motion priors of Stable Video Diffusion (SVD) and incorporates task-specific information through a unified face restoration framework. A learnable task embedding is introduced to enhance task identification. Meanwhile, a novel Unified Latent Regularization (ULR) is employed to encourage the shared feature representation learning among different subtasks. To further enhance the restoration quality and temporal stability, we introduce the facial prior learning and the self-referred refinement as auxiliary strategies used for both training and inference. The proposed framework effectively combines the complementary strengths of these tasks, enhancing temporal coherence and achieving superior restoration quality. This work advances the state-of-the-art in video FR and establishes a new paradigm for generalized video face restoration.
Authors: Junjie Yang, Haibo Wan, Zhihai Shang
Abstract: Remote sensing (RS) change detection incurs a high cost because of false negatives, which are more costly than false positives. Existing frameworks, struggling to improve the Precision metric to reduce the cost of false positive, still have limitations in focusing on the change of interest, which leads to missed detections and discontinuity issues. This work tackles these issues by enhancing feature learning capabilities and integrating the frequency components of feature information, with a strategy to incrementally boost the Recall value. We propose an enhanced hybrid of CNN and Transformer network (EHCTNet) for effectively mining the change information of interest. Firstly, a dual branch feature extraction module is used to extract the multi scale features of RS images. Secondly, the frequency component of these features is exploited by a refined module I. Thirdly, an enhanced token mining module based on the Kolmogorov Arnold Network is utilized to derive semantic information. Finally, the semantic change information's frequency component, beneficial for final detection, is mined from the refined module II. Extensive experiments validate the effectiveness of EHCTNet in comprehending complex changes of interest. The visualization outcomes show that EHCTNet detects more intact and continuous changed areas and perceives more accurate neighboring distinction than state of the art models.
Authors: Xiyuan Gao, Bing Cao, Pengfei Zhu, Nannan Wang, Qinghua Hu
Abstract: The strength of multimodal learning lies in its ability to integrate information from various sources, providing rich and comprehensive insights. However, in real-world scenarios, multi-modal systems often face the challenge of dynamic modality contributions, the dominance of different modalities may change with the environments, leading to suboptimal performance in multimodal learning. Current methods mainly enhance weak modalities to balance multimodal representation bias, which inevitably optimizes from a partialmodality perspective, easily leading to performance descending for dominant modalities. To address this problem, we propose an Asymmetric Reinforcing method against Multimodal representation bias (ARM). Our ARM dynamically reinforces the weak modalities while maintaining the ability to represent dominant modalities through conditional mutual information. Moreover, we provide an in-depth analysis that optimizing certain modalities could cause information loss and prevent leveraging the full advantages of multimodal data. By exploring the dominance and narrowing the contribution gaps between modalities, we have significantly improved the performance of multimodal learning, making notable progress in mitigating imbalanced multimodal learning.
Authors: Lixiong Qin, Shilong Ou, Miaoxuan Zhang, Jiangning Wei, Yuhang Zhang, Xiaoshuai Song, Yuchen Liu, Mei Wang, Weiran Xu
Abstract: Faces and humans are crucial elements in social interaction and are widely included in everyday photos and videos. Therefore, a deep understanding of faces and humans will enable multi-modal assistants to achieve improved response quality and broadened application scope. Currently, the multi-modal assistant community lacks a comprehensive and scientific evaluation of face and human understanding abilities. In this paper, we first propose a hierarchical ability taxonomy that includes three levels of abilities. Then, based on this taxonomy, we collect images and annotations from publicly available datasets in the face and human community and build a semi-automatic data pipeline to produce problems for the new benchmark. Finally, the obtained Face-Human-Bench comprises a development set with 900 problems and a test set with 1800 problems, supporting both English and Chinese. We conduct evaluations over 25 mainstream multi-modal large language models (MLLMs) with our Face-Human-Bench, focusing on the correlation between abilities, the impact of the relative position of targets on performance, and the impact of Chain of Thought (CoT) prompting on performance. Moreover, inspired by multi-modal agents, we also explore which abilities of MLLMs need to be supplemented by specialist models.
Authors: Yongle Huang, Haodong Chen, Zhenbang Xu, Zihan Jia, Haozhou Sun, Dian Shao
Abstract: Human action understanding is crucial for the advancement of multimodal systems. While recent developments, driven by powerful large language models (LLMs), aim to be general enough to cover a wide range of categories, they often overlook the need for more specific capabilities. In this work, we address the more challenging task of Fine-grained Action Recognition (FAR), which focuses on detailed semantic labels within shorter temporal duration (e.g., "salto backward tucked with 1 turn"). Given the high costs of annotating fine-grained labels and the substantial data needed for fine-tuning LLMs, we propose to adopt semi-supervised learning (SSL). Our framework, SeFAR, incorporates several innovative designs to tackle these challenges. Specifically, to capture sufficient visual details, we construct Dual-level temporal elements as more effective representations, based on which we design a new strong augmentation strategy for the Teacher-Student learning paradigm through involving moderate temporal perturbation. Furthermore, to handle the high uncertainty within the teacher model's predictions for FAR, we propose the Adaptive Regulation to stabilize the learning process. Experiments show that SeFAR achieves state-of-the-art performance on two FAR datasets, FineGym and FineDiving, across various data scopes. It also outperforms other semi-supervised methods on two classical coarse-grained datasets, UCF101 and HMDB51. Further analysis and ablation studies validate the effectiveness of our designs. Additionally, we show that the features extracted by our SeFAR could largely promote the ability of multimodal foundation models to understand fine-grained and domain-specific semantics.
Authors: Mengjie Qin, Yuchao Feng, Zongliang Wu, Yulun Zhang, Xin Yuan
Abstract: In the coded aperture snapshot spectral imaging system, Deep Unfolding Networks (DUNs) have made impressive progress in recovering 3D hyperspectral images (HSIs) from a single 2D measurement. However, the inherent nonlinear and ill-posed characteristics of HSI reconstruction still pose challenges to existing methods in terms of accuracy and stability. To address this issue, we propose a Mamba-inspired Joint Unfolding Network (MiJUN), which integrates physics-embedded DUNs with learning-based HSI imaging. Firstly, leveraging the concept of trapezoid discretization to expand the representation space of unfolding networks, we introduce an accelerated unfolding network scheme. This approach can be interpreted as a generalized accelerated half-quadratic splitting with a second-order differential equation, which reduces the reliance on initial optimization stages and addresses challenges related to long-range interactions. Crucially, within the Mamba framework, we restructure the Mamba-inspired global-to-local attention mechanism by incorporating a selective state space model and an attention mechanism. This effectively reinterprets Mamba as a variant of the Transformer} architecture, improving its adaptability and efficiency. Furthermore, we refine the scanning strategy with Mamba by integrating the tensor mode-$k$ unfolding into the Mamba network. This approach emphasizes the low-rank properties of tensors along various modes, while conveniently facilitating 12 scanning directions. Numerical and visual comparisons on both simulation and real datasets demonstrate the superiority of our proposed MiJUN, and achieving overwhelming detail representation.
Authors: Leandro Di Bella, Yangxintong Lyu, Bruno Cornelis, Adrian Munteanu
Abstract: The evolution of Advanced Driver Assistance Systems (ADAS) has increased the need for robust and generalizable algorithms for multi-object tracking. Traditional statistical model-based tracking methods rely on predefined motion models and assumptions about system noise distributions. Although computationally efficient, they often lack adaptability to varying traffic scenarios and require extensive manual design and parameter tuning. To address these issues, we propose a novel 3D multi-object tracking approach for vehicles, HybridTrack, which integrates a data-driven Kalman Filter (KF) within a tracking-by-detection paradigm. In particular, it learns the transition residual and Kalman gain directly from data, which eliminates the need for manual motion and stochastic parameter modeling. Validated on the real-world KITTI dataset, HybridTrack achieves 82.08% HOTA accuracy, significantly outperforming state-of-the-art methods. We also evaluate our method under different configurations, achieving the fastest processing speed of 112 FPS. Consequently, HybridTrack eliminates the dependency on scene-specific designs while improving performance and maintaining real-time efficiency. The code will be publicly available at the time of publishing: https://github.com/leandro-svg/HybridTrack.git.
Authors: Bohang Sun, Pietro Li\`o
Abstract: In this study, we introduce the Multi-Head Explainer (MHEX), a versatile and modular framework that enhances both the explainability and accuracy of Convolutional Neural Networks (CNNs) and Transformer-based models. MHEX consists of three core components: an Attention Gate that dynamically highlights task-relevant features, Deep Supervision that guides early layers to capture fine-grained details pertinent to the target class, and an Equivalent Matrix that unifies refined local and global representations to generate comprehensive saliency maps. Our approach demonstrates superior compatibility, enabling effortless integration into existing residual networks like ResNet and Transformer architectures such as BERT with minimal modifications. Extensive experiments on benchmark datasets in medical imaging and text classification show that MHEX not only improves classification accuracy but also produces highly interpretable and detailed saliency scores.
Authors: Jianyi Wang, Zhijie Lin, Meng Wei, Yang Zhao, Ceyuan Yang, Chen Change Loy, Lu Jiang
Abstract: Video restoration poses non-trivial challenges in maintaining fidelity while recovering temporally consistent details from unknown degradations in the wild. Despite recent advances in diffusion-based restoration, these methods often face limitations in generation capability and sampling efficiency. In this work, we present SeedVR, a diffusion transformer designed to handle real-world video restoration with arbitrary length and resolution. The core design of SeedVR lies in the shifted window attention that facilitates effective restoration on long video sequences. SeedVR further supports variable-sized windows near the boundary of both spatial and temporal dimensions, overcoming the resolution constraints of traditional window attention. Equipped with contemporary practices, including causal video autoencoder, mixed image and video training, and progressive training, SeedVR achieves highly-competitive performance on both synthetic and real-world benchmarks, as well as AI-generated videos. Extensive experiments demonstrate SeedVR's superiority over existing methods for generic video restoration.
Authors: Dong Shu, Haiyan Zhao, Jingyu Hu, Weiru Liu, Lu Cheng, Mengnan Du
Abstract: Large Vision-Language Models (LVLMs) have demonstrated remarkable capabilities in processing both visual and textual information. However, the critical challenge of alignment between visual and linguistic representations is not fully understood. This survey presents a comprehensive examination of alignment and misalignment in LVLMs through an explainability lens. We first examine the fundamentals of alignment, exploring its representational and behavioral aspects, training methodologies, and theoretical foundations. We then analyze misalignment phenomena across three semantic levels: object, attribute, and relational misalignment. Our investigation reveals that misalignment emerges from challenges at multiple levels: the data level, the model level, and the inference level. We provide a comprehensive review of existing mitigation strategies, categorizing them into parameter-frozen and parameter-tuning approaches. Finally, we outline promising future research directions, emphasizing the need for standardized evaluation protocols and in-depth explainability studies.
Authors: Austin T. Wang, ZeMing Gong, Angel X. Chang
Abstract: 3D visual grounding (3DVG) involves localizing entities in a 3D scene referred to by natural language text. Such models are useful for embodied AI and scene retrieval applications, which involve searching for objects or patterns using natural language descriptions. While recent works have focused on LLM-based scaling of 3DVG datasets, these datasets do not capture the full range of potential prompts which could be specified in the English language. To ensure that we are scaling up and testing against a useful and representative set of prompts, we propose a framework for linguistically analyzing 3DVG prompts and introduce Visual Grounding with Diverse Language in 3D (ViGiL3D), a diagnostic dataset for evaluating visual grounding methods against a diverse set of language patterns. We evaluate existing open-vocabulary 3DVG methods to demonstrate that these methods are not yet proficient in understanding and identifying the targets of more challenging, out-of-distribution prompts, toward real-world applications.
Authors: Z. Zhang, B. Liu, J. Bao, L. Chen, S. Zhu, J. Yu
Abstract: Recent text-to-image generation favors various forms of spatial conditions, e.g., masks, bounding boxes, and key points. However, the majority of the prior art requires form-specific annotations to fine-tune the original model, leading to poor test-time generalizability. Meanwhile, existing training-free methods work well only with simplified prompts and spatial conditions. In this work, we propose a novel yet generic test-time controllable generation method that aims at natural text prompts and complex conditions. Specifically, we decouple spatial conditions into semantic and geometric conditions and then enforce their consistency during the image-generation process individually. As for the former, we target bridging the gap between the semantic condition and text prompts, as well as the gap between such condition and the attention map from diffusion models. To achieve this, we propose to first complete the prompt w.r.t. semantic condition, and then remove the negative impact of distracting prompt words by measuring their statistics in attention maps as well as distances in word space w.r.t. this condition. To further cope with the complex geometric conditions, we introduce a geometric transform module, in which Region-of-Interests will be identified in attention maps and further used to translate category-wise latents w.r.t. geometric condition. More importantly, we propose a diffusion-based latents-refill method to explicitly remove the impact of latents at the RoI, reducing the artifacts on generated images. Experiments on Coco-stuff dataset showcase 30$\%$ relative boost compared to SOTA training-free methods on layout consistency evaluation metrics.
Authors: Ben Vardi, Oron Nir, Ariel Shamir
Abstract: Recent Vision-Language Models (VLMs) have demonstrated remarkable capabilities in visual understanding and reasoning, and in particular on multiple-choice Visual Question Answering (VQA). Still, these models can make distinctly unnatural errors, for example, providing (wrong) answers to unanswerable VQA questions, such as questions asking about objects that do not appear in the image. To address this issue, we propose CLIP-UP: CLIP-based Unanswerable Problem detection, a novel lightweight method for equipping VLMs with the ability to withhold answers to unanswerable questions. By leveraging CLIP to extract question-image alignment information, CLIP-UP requires only efficient training of a few additional layers, while keeping the original VLMs' weights unchanged. Tested across LLaVA models, CLIP-UP achieves state-of-the-art results on the MM-UPD benchmark for assessing unanswerability in multiple-choice VQA, while preserving the original performance on other tasks.
Authors: Rasel Ahmed Bhuiyan, Mateusz Trokielewicz, Piotr Maciejewicz, Sherri Bucher, Adam Czajka
Abstract: Non-invasive, efficient, physical token-less, accurate and stable identification methods for newborns may prevent baby swapping at birth, limit baby abductions and improve post-natal health monitoring across geographies, within the context of both the formal (i.e., hospitals) and informal (i.e., humanitarian and fragile settings) health sectors. This paper explores the feasibility of application iris recognition to build biometric identifiers for 4-6 week old infants. We (a) collected near infrared (NIR) iris images from 17 infants using a specially-designed NIR iris sensor; (b) evaluated six iris recognition methods to assess readiness of the state-of-the-art iris recognition to be applied to newborns and infants; (c) proposed a new segmentation model that correctly detects iris texture within infants iris images, and coupled it with several iris texture encoding approaches to offer, to the first of our knowledge, a fully-operational infant iris recognition system; and, (d) trained a StyleGAN-based model to synthesize iris images mimicking samples acquired from infants to deliver to the research community privacy-safe infant iris images. The proposed system, incorporating the specially-designed iris sensor and segmenter, and applied to the collected infant iris samples, achieved Equal Error Rate (EER) of 3\% and Area Under ROC Curve (AUC) of 99\%, compared to EER$\geq$20\% and AUC$\leq$88\% obtained for state of the art adult iris recognition systems. This suggests that it may be feasible to design methods that succesfully extract biometric features from infant irises.
Authors: Yidi Shao, Chen Change Loy, Bo Dai
Abstract: Garment animation is ubiquitous in various applications, such as virtual reality, gaming, and film producing. Recently, learning-based approaches obtain compelling performance in animating diverse garments under versatile scenarios. Nevertheless, to mimic the deformations of the observed garments, data-driven methods require large scale of garment data, which are both resource-wise expensive and time-consuming. In addition, forcing models to match the dynamics of observed garment animation may hinder the potentials to generalize to unseen cases. In this paper, instead of using garment-wise supervised-learning we adopt a disentangled scheme to learn how to animate observed garments: 1). learning constitutive behaviors from the observed cloth; 2). dynamically animate various garments constrained by the learned constitutive laws. Specifically, we propose Energy Unit network (EUNet) to model the constitutive relations in the format of energy. Without the priors from analytical physics models and differentiable simulation engines, EUNet is able to directly capture the constitutive behaviors from the observed piece of cloth and uniformly describes the change of energy caused by deformations, such as stretching and bending. We further apply the pre-trained EUNet to animate various garments based on energy optimizations. The disentangled scheme alleviates the need of garment data and enables us to utilize the dynamics of a piece of cloth for animating garments. Experiments show that while EUNet effectively delivers the energy gradients due to the deformations, models constrained by EUNet achieve more stable and physically plausible performance comparing with those trained in garment-wise supervised manner. Code is available at https://github.com/ftbabi/EUNet_NeurIPS2024.git .
Authors: Haixu Liu, Zerui Tao, Wenzhen Dong, Qiuzhuang Sun
Abstract: This paper provides a novel 3D medical image segmentation model structure called nnY-Net. This name comes from the fact that our model adds a cross-attention module at the bottom of the U-net structure to form a Y structure. We integrate the advantages of the two latest SOTA models, MedNeXt and SwinUNETR, and use Swin Transformer as the encoder and ConvNeXt as the decoder to innovatively design the Swin-NeXt structure. Our model uses the lowest-level feature map of the encoder as Key and Value and uses patient features such as pathology and treatment information as Query to calculate the attention weights in a Cross Attention module. Moreover, we simplify some pre- and post-processing as well as data enhancement methods in 3D image segmentation based on the dynUnet and nnU-net frameworks. We integrate our proposed Swin-NeXt with Cross-Attention framework into this framework. Last, we construct a DiceFocalCELoss to improve the training efficiency for the uneven data convergence of voxel classification.
Authors: Or Patashnik, Rinon Gal, Daniil Ostashev, Sergey Tulyakov, Kfir Aberman, Daniel Cohen-Or
Abstract: Personalizing text-to-image models to generate images of specific subjects across diverse scenes and styles is a rapidly advancing field. Current approaches often face challenges in maintaining a balance between identity preservation and alignment with the input text prompt. Some methods rely on a single textual token to represent a subject, which limits expressiveness, while others employ richer representations but disrupt the model's prior, diminishing prompt alignment. In this work, we introduce Nested Attention, a novel mechanism that injects a rich and expressive image representation into the model's existing cross-attention layers. Our key idea is to generate query-dependent subject values, derived from nested attention layers that learn to select relevant subject features for each region in the generated image. We integrate these nested layers into an encoder-based personalization method, and show that they enable high identity preservation while adhering to input text prompts. Our approach is general and can be trained on various domains. Additionally, its prior preservation allows us to combine multiple personalized subjects from different domains in a single image.
Authors: Chun-Hao Paul Huang, Jae Shin Yoon, Hyeonho Jeong, Niloy Mitra, Duygu Ceylan
Abstract: Inspired by the emergent 3D capabilities in image generators, we explore whether video generators similarly exhibit 3D awareness. Using structure-from-motion (SfM) as a benchmark for 3D tasks, we investigate if intermediate features from OpenSora, a video generation model, can support camera pose estimation. We first examine native 3D awareness in video generation features by routing raw intermediate outputs to SfM-prediction modules like DUSt3R. Then, we explore the impact of fine-tuning on camera pose estimation to enhance 3D awareness. Results indicate that while video generator features have limited inherent 3D awareness, task-specific supervision significantly boosts their accuracy for camera pose estimation, resulting in competitive performance. The proposed unified model, named JOG3R, produces camera pose estimates with competitive quality without degrading video generation quality.
Authors: Yaxian Wang, Henghui Ding, Shuting He, Xudong Jiang, Bifan Wei, Jun Liu
Abstract: In this work, we address the challenging task of Generalized Referring Expression Comprehension (GREC). Compared to the classic Referring Expression Comprehension (REC) that focuses on single-target expressions, GREC extends the scope to a more practical setting by further encompassing no-target and multi-target expressions. Existing REC methods face challenges in handling the complex cases encountered in GREC, primarily due to their fixed output and limitations in multi-modal representations. To address these issues, we propose a Hierarchical Alignment-enhanced Adaptive Grounding Network (HieA2G) for GREC, which can flexibly deal with various types of referring expressions. First, a Hierarchical Multi-modal Semantic Alignment (HMSA) module is proposed to incorporate three levels of alignments, including word-object, phrase-object, and text-image alignment. It enables hierarchical cross-modal interactions across multiple levels to achieve comprehensive and robust multi-modal understanding, greatly enhancing grounding ability for complex cases. Then, to address the varying number of target objects in GREC, we introduce an Adaptive Grounding Counter (AGC) to dynamically determine the number of output targets. Additionally, an auxiliary contrastive loss is employed in AGC to enhance object-counting ability by pulling in multi-modal features with the same counting and pushing away those with different counting. Extensive experimental results show that HieA2G achieves new state-of-the-art performance on the challenging GREC task and also the other 4 tasks, including REC, Phrase Grounding, Referring Expression Segmentation (RES), and Generalized Referring Expression Segmentation (GRES), demonstrating the remarkable superiority and generalizability of the proposed HieA2G.
Authors: Yoshitomo Matsubara, Matteo Mendula, Marco Levorato
Abstract: Split computing ($\neq$ split learning) is a promising approach to deep learning models for resource-constrained edge computing systems, where weak sensor (mobile) devices are wirelessly connected to stronger edge servers through channels with limited communication capacity. State-of-theart work on split computing presents methods for single tasks such as image classification, object detection, or semantic segmentation. The application of existing methods to multitask problems degrades model accuracy and/or significantly increase runtime latency. In this study, we propose Ladon, the first multi-task-head supervised compression model for multi-task split computing. Experimental results show that the multi-task supervised compression model either outperformed or rivaled strong lightweight baseline models in terms of predictive performance for ILSVRC 2012, COCO 2017, and PASCAL VOC 2012 datasets while learning compressed representations at its early layers. Furthermore, our models reduced end-to-end latency (by up to 95.4%) and energy consumption of mobile devices (by up to 88.2%) in multi-task split computing scenarios.
Authors: Xudong Jiang, Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys
Abstract: Learning-based visual localization methods that use scene coordinate regression (SCR) offer the advantage of smaller map sizes. However, on datasets with complex illumination changes or image-level ambiguities, it remains a less robust alternative to feature matching methods. This work aims to close the gap. We introduce a covisibility graph-based global encoding learning and data augmentation strategy, along with a depth-adjusted reprojection loss to facilitate implicit triangulation. Additionally, we revisit the network architecture and local feature extraction module. Our method achieves state-of-the-art on challenging large-scale datasets without relying on network ensembles or 3D supervision. On Aachen Day-Night, we are 10$\times$ more accurate than previous SCR methods with similar map sizes and require at least 5$\times$ smaller map sizes than any other SCR method while still delivering superior accuracy. Code will be available at: https://github.com/cvg/scrstudio .
Authors: Haixu Liu, Wenning Wang, Haoxiang Zheng, Penghao Jiang, Qirui Wang, Ruiqing Yan, Qiuzhuang Sun
Abstract: This work aims to predict the popularity of short videos using the videos themselves and their related features. Popularity is measured by four key engagement metrics: view count, like count, comment count, and share count. This study employs video classification models with different architectures and training methods as backbone networks to extract video modality features. Meanwhile, the cleaned video captions are incorporated into a carefully designed prompt framework, along with the video, as input for video-to-text generation models, which generate detailed text-based video content understanding. These texts are then encoded into vectors using a pre-trained BERT model. Based on the six sets of vectors mentioned above, a neural network is trained for each of the four prediction metrics. Moreover, the study conducts data mining and feature engineering based on the video and tabular data, constructing practical features such as the total frequency of hashtag appearances, the total frequency of mention appearances, video duration, frame count, frame rate, and total time online. Multiple machine learning models are trained, and the most stable model, XGBoost, is selected. Finally, the predictions from the neural network and XGBoost models are averaged to obtain the final result.
Authors: Jingfeng Yao, Xinggang Wang
Abstract: Latent diffusion models with Transformer architectures excel at generating high-fidelity images. However, recent studies reveal an optimization dilemma in this two-stage design: while increasing the per-token feature dimension in visual tokenizers improves reconstruction quality, it requires substantially larger diffusion models and more training iterations to achieve comparable generation performance. Consequently, existing systems often settle for sub-optimal solutions, either producing visual artifacts due to information loss within tokenizers or failing to converge fully due to expensive computation costs. We argue that this dilemma stems from the inherent difficulty in learning unconstrained high-dimensional latent spaces. To address this, we propose aligning the latent space with pre-trained vision foundation models when training the visual tokenizers. Our proposed VA-VAE (Vision foundation model Aligned Variational AutoEncoder) significantly expands the reconstruction-generation frontier of latent diffusion models, enabling faster convergence of Diffusion Transformers (DiT) in high-dimensional latent spaces. To exploit the full potential of VA-VAE, we build an enhanced DiT baseline with improved training strategies and architecture designs, termed LightningDiT. The integrated system achieves state-of-the-art (SOTA) performance on ImageNet 256x256 generation with an FID score of 1.35 while demonstrating remarkable training efficiency by reaching an FID score of 2.11 in just 64 epochs--representing an over 21 times convergence speedup compared to the original DiT. Models and codes are available at: https://github.com/hustvl/LightningDiT.
Authors: Gaurav Parmar, Or Patashnik, Kuan-Chieh Wang, Daniil Ostashev, Srinivasa Narasimhan, Jun-Yan Zhu, Daniel Cohen-Or, Kfir Aberman
Abstract: We introduce a method for composing object-level visual prompts within a text-to-image diffusion model. Our approach addresses the task of generating semantically coherent compositions across diverse scenes and styles, similar to the versatility and expressiveness offered by text prompts. A key challenge in this task is to preserve the identity of the objects depicted in the input visual prompts, while also generating diverse compositions across different images. To address this challenge, we introduce a new KV-mixed cross-attention mechanism, in which keys and values are learned from distinct visual representations. The keys are derived from an encoder with a small bottleneck for layout control, whereas the values come from a larger bottleneck encoder that captures fine-grained appearance details. By mixing keys and values from these complementary sources, our model preserves the identity of the visual prompts while supporting flexible variations in object arrangement, pose, and composition. During inference, we further propose object-level compositional guidance to improve the method's identity preservation and layout correctness. Results show that our technique produces diverse scene compositions that preserve the unique characteristics of each visual prompt, expanding the creative potential of text-to-image generation.
Authors: Xincheng Shuai, Henghui Ding, Zhenyuan Qin, Hao Luo, Xingjun Ma, Dacheng Tao
Abstract: Controlling the movements of dynamic objects and the camera within generated videos is a meaningful yet challenging task. Due to the lack of datasets with comprehensive motion annotations, existing algorithms can not simultaneously control the motions of both camera and objects, resulting in limited controllability over generated contents. To address this issue and facilitate the research in this field, we introduce a Synthetic Dataset for Free-Form Motion Control (SynFMC). The proposed SynFMC dataset includes diverse objects and environments and covers various motion patterns according to specific rules, simulating common and complex real-world scenarios. The complete 6D pose information facilitates models learning to disentangle the motion effects from objects and the camera in a video. To validate the effectiveness and generalization of SynFMC, we further propose a method, Free-Form Motion Control (FMC). FMC enables independent or simultaneous control of object and camera movements, producing high-fidelity videos. Moreover, it is compatible with various personalized text-to-image (T2I) models for different content styles. Extensive experiments demonstrate that the proposed FMC outperforms previous methods across multiple scenarios.
Authors: Jihoon Chung, Tyler Zhu, Max Gonzalez Saez-Diez, Juan Carlos Niebles, Honglu Zhou, Olga Russakovsky
Abstract: The recent advent of Large Language Models (LLMs) has ushered sophisticated reasoning capabilities into the realm of video through Video Large Language Models (VideoLLMs). However, VideoLLMs currently rely on a single vision encoder for all of their visual processing, which limits the amount and type of visual information that can be conveyed to the LLM. Our method, MERV, Multi-Encoder Representation of Videos, instead leverages multiple frozen visual encoders to create a unified representation of a video, providing the VideoLLM with a comprehensive set of specialized visual knowledge. Spatio-temporally aligning the features from each encoder allows us to tackle a wider range of open-ended and multiple-choice video understanding questions and outperform prior state-of-the-art works. MERV is up to 3.7% better in accuracy than Video-LLaVA across the standard suite video understanding benchmarks, while also having a better Video-ChatGPT score. We also improve upon SeViLA, the previous best on zero-shot Perception Test accuracy, by 2.2%. MERV introduces minimal extra parameters and trains faster than equivalent single-encoder methods while parallelizing the visual processing. Finally, we provide qualitative evidence that MERV successfully captures domain knowledge from each of its encoders. Our results offer promising directions in utilizing multiple vision encoders for comprehensive video understanding.
Authors: Yuanpeng Tu, Hao Luo, Xi Chen, Sihui Ji, Xiang Bai, Hengshuang Zhao
Abstract: Despite significant advancements in video generation, inserting a given object into videos remains a challenging task. The difficulty lies in preserving the appearance details of the reference object and accurately modeling coherent motions at the same time. In this paper, we propose VideoAnydoor, a zero-shot video object insertion framework with high-fidelity detail preservation and precise motion control. Starting from a text-to-video model, we utilize an ID extractor to inject the global identity and leverage a box sequence to control the overall motion. To preserve the detailed appearance and meanwhile support fine-grained motion control, we design a pixel warper. It takes the reference image with arbitrary key-points and the corresponding key-point trajectories as inputs. It warps the pixel details according to the trajectories and fuses the warped features with the diffusion U-Net, thus improving detail preservation and supporting users in manipulating the motion trajectories. In addition, we propose a training strategy involving both videos and static images with a reweight reconstruction loss to enhance insertion quality. VideoAnydoor demonstrates significant superiority over existing methods and naturally supports various downstream applications (e.g., talking head generation, video virtual try-on, multi-region editing) without task-specific fine-tuning.
Authors: Zhangyang Qi, Zhixiong Zhang, Ye Fang, Jiaqi Wang, Hengshuang Zhao
Abstract: In recent years, 2D Vision-Language Models (VLMs) have made significant strides in image-text understanding tasks. However, their performance in 3D spatial comprehension, which is critical for embodied intelligence, remains limited. Recent advances have leveraged 3D point clouds and multi-view images as inputs, yielding promising results. However, we propose exploring a purely vision-based solution inspired by human perception, which merely relies on visual cues for 3D spatial understanding. This paper empirically investigates the limitations of VLMs in 3D spatial knowledge, revealing that their primary shortcoming lies in the lack of global-local correspondence between the scene and individual frames. To address this, we introduce GPT4Scene, a novel visual prompting paradigm in VLM training and inference that helps build the global-local relationship, significantly improving the 3D spatial understanding of indoor scenes. Specifically, GPT4Scene constructs a 3D Bird's Eye View (BEV) image from the video and marks consistent object IDs across both frames and the BEV image. The model then inputs the concatenated BEV image and video frames with markers. In zero-shot evaluations, GPT4Scene improves performance over closed-source VLMs like GPT-4o. Additionally, we prepare a processed video dataset consisting of 165K text annotation to fine-tune open-source VLMs, achieving state-of-the-art performance on all 3D understanding tasks. Surprisingly, after training with the GPT4Scene paradigm, VLMs consistently improve during inference, even without visual prompting and BEV image as explicit correspondence. It demonstrates that the proposed paradigm helps VLMs develop an intrinsic ability to understand 3D scenes, which paves the way for a noninvasive approach to extending pre-trained VLMs for 3D scene understanding.
Authors: Witold Wydma\'nski, Ulvi Movsum-zada, Jacek Tabor, Marek \'Smieja
Abstract: Although deep learning models have had great success in natural language processing and computer vision, we do not observe comparable improvements in the case of tabular data, which is still the most common data type used in biological, industrial and financial applications. In particular, it is challenging to transfer large-scale pre-trained models to downstream tasks defined on small tabular datasets. To address this, we propose VisTabNet -- a cross-modal transfer learning method, which allows for adapting Vision Transformer (ViT) with pre-trained weights to process tabular data. By projecting tabular inputs to patch embeddings acceptable by ViT, we can directly apply a pre-trained Transformer Encoder to tabular inputs. This approach eliminates the conceptual cost of designing a suitable architecture for processing tabular data, while reducing the computational cost of training the model from scratch. Experimental results on multiple small tabular datasets (less than 1k samples) demonstrate VisTabNet's superiority, outperforming both traditional ensemble methods and recent deep learning models. The proposed method goes beyond conventional transfer learning practice and shows that pre-trained image models can be transferred to solve tabular problems, extending the boundaries of transfer learning.
Authors: Amirhossein Nadiri, Jing Li, Ali Faraji, Ghadeer Abuoda, Manos Papagelis
Abstract: Trajectory prediction aims to estimate an entity's future path using its current position and historical movement data, benefiting fields like autonomous navigation, robotics, and human movement analytics. Deep learning approaches have become key in this area, utilizing large-scale trajectory datasets to model movement patterns, but face challenges in managing complex spatial dependencies and adapting to dynamic environments. To address these challenges, we introduce TrajLearn, a novel model for trajectory prediction that leverages generative modeling of higher-order mobility flows based on hexagonal spatial representation. TrajLearn predicts the next $k$ steps by integrating a customized beam search for exploring multiple potential paths while maintaining spatial continuity. We conducted a rigorous evaluation of TrajLearn, benchmarking it against leading state-of-the-art approaches and meaningful baselines. The results indicate that TrajLearn achieves significant performance gains, with improvements of up to ~40% across multiple real-world trajectory datasets. In addition, we evaluated different prediction horizons (i.e., various values of $k$), conducted resolution sensitivity analysis, and performed ablation studies to assess the impact of key model components. Furthermore, we developed a novel algorithm to generate mixed-resolution maps by hierarchically subdividing hexagonal regions into finer segments within a specified observation area. This approach supports selective detailing, applying finer resolution to areas of interest or high activity (e.g., urban centers) while using coarser resolution for less significant regions (e.g., rural areas), effectively reducing data storage requirements and computational overhead. We promote reproducibility and adaptability by offering complete code, data, and detailed documentation with flexible configuration options for various applications.
Authors: Rajat Talak, Charis Georgiou, Jingnan Shi, Luca Carlone
Abstract: Robust training of machine learning models in the presence of outliers has garnered attention across various domains. The use of robust losses is a popular approach and is known to mitigate the impact of outliers. We bring to light two literatures that have diverged in their ways of designing robust losses: one using M-estimation, which is popular in robotics and computer vision, and another using a risk-minimization framework, which is popular in deep learning. We first show that a simple modification of the Black-Rangarajan duality provides a unifying view. The modified duality brings out a definition of a robust loss kernel $\sigma$ that is satisfied by robust losses in both the literatures. Secondly, using the modified duality, we propose an Adaptive Alternation Algorithm (AAA) for training machine learning models with outliers. The algorithm iteratively trains the model by using a weighted version of the non-robust loss, while updating the weights at each iteration. The algorithm is augmented with a novel parameter update rule by interpreting the weights as inlier probabilities, and obviates the need for complex parameter tuning. Thirdly, we investigate convergence of the adaptive alternation algorithm to outlier-free optima. Considering arbitrary outliers (i.e., with no distributional assumption on the outliers), we show that the use of robust loss kernels {\sigma} increases the region of convergence. We experimentally show the efficacy of our algorithm on regression, classification, and neural scene reconstruction problems. We release our implementation code: https://github.com/MIT-SPARK/ORT.
Authors: Ashay Athalye, Nishanth Kumar, Tom Silver, Yichao Liang, Tom\'as Lozano-P\'erez, Leslie Pack Kaelbling
Abstract: Our aim is to learn to solve long-horizon decision-making problems in highly-variable, combinatorially-complex robotics domains given raw sensor input in the form of images. Previous work has shown that one way to achieve this aim is to learn a structured abstract transition model in the form of symbolic predicates and operators, and then plan within this model to solve novel tasks at test time. However, these learned models do not ground directly into pixels from just a handful of demonstrations. In this work, we propose to invent predicates that operate directly over input images by leveraging the capabilities of pretrained vision-language models (VLMs). Our key idea is that, given a set of demonstrations, a VLM can be used to propose a set of predicates that are potentially relevant for decision-making and then to determine the truth values of these predicates in both the given demonstrations and new image inputs. We build upon an existing framework for predicate invention, which generates feature-based predicates operating on object-centric states, to also generate visual predicates that operate on images. Experimentally, we show that our approach -- pix2pred -- is able to invent semantically meaningful predicates that enable generalization to novel, complex, and long-horizon tasks across two simulated robotic environments.
Authors: Wenhao Dong, Yueyang Li, Weiming Zeng, Lei Chen, Hongjie Yan, Wai Ting Siok, Nizhuan Wang
Abstract: Many existing methods that use functional magnetic resonance imaging (fMRI) classify brain disorders, such as autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD), often overlook the integration of spatial and temporal dependencies of the blood oxygen level-dependent (BOLD) signals, which may lead to inaccurate or imprecise classification results. To solve this problem, we propose a Spatio-Temporal Aggregation eorganization ransformer (STARFormer) that effectively captures both spatial and temporal features of BOLD signals by incorporating three key modules. The region of interest (ROI) spatial structure analysis module uses eigenvector centrality (EC) to reorganize brain regions based on effective connectivity, highlighting critical spatial relationships relevant to the brain disorder. The temporal feature reorganization module systematically segments the time series into equal-dimensional window tokens and captures multiscale features through variable window and cross-window attention. The spatio-temporal feature fusion module employs a parallel transformer architecture with dedicated temporal and spatial branches to extract integrated features. The proposed STARFormer has been rigorously evaluated on two publicly available datasets for the classification of ASD and ADHD. The experimental results confirm that the STARFormer achieves state-of-the-art performance across multiple evaluation metrics, providing a more accurate and reliable tool for the diagnosis of brain disorders and biomedical research. The codes will be available at: https://github.com/NZWANG/STARFormer.
Authors: Zhenhan Huang, Tejaswini Pedapati, Pin-Yu Chen, Jianxi Gao
Abstract: Prompt learning is an effective way to exploit the potential of large-scale pre-trained foundational models. Continuous prompts parameterize context tokens in prompts by turning them into differentiable vectors. Deep continuous prompts insert prompts not only in the input but also in the intermediate hidden representations. Manually designed deep continuous prompts exhibit a remarkable improvement compared to the zero-shot pre-trained model on downstream tasks. How to automate the continuous prompt design is an underexplored area, and a fundamental question arises, is manually designed deep prompt strategy optimal? To answer this question, we propose a method dubbed differentiable prompt learning (DPL). The DPL method is formulated as an optimization problem to automatically determine the optimal context length of the prompt to be added to each layer, where the objective is to maximize the performance. We test the DPL method on the pre-trained CLIP. We empirically find that by using only limited data, our DPL method can find deep continuous prompt configuration with high confidence. The performance on the downstream tasks exhibits the superiority of the automatic design: our method boosts the average test accuracy by 2.60% on 11 datasets compared to baseline methods. Besides, our method focuses only on the prompt configuration (i.e. context length for each layer), which means that our method is compatible with the baseline methods that have sophisticated designs to boost the performance. The DPL method can be deployed to large language models or computer vision models at no cost.
Authors: Lu Zhang, Liang Zeng
Abstract: The proliferation of AI-generated images necessitates effective watermarking to protect intellectual property and identify fake content. While existing training-based watermarking methods show promise, they often struggle with generalization across diverse prompts and tend to produce noticeable artifacts. To this end, we introduce a provably generalizable image watermarking method for Latent Diffusion Models with Self-Augmented Training (SAT-LDM), which aligns the training and testing phases by a free generation distribution to bolster the watermarking module's generalization capabilities. We theoretically consolidate our method by proving that the free generation distribution contributes to its tight generalization bound without the need to collect new data. Extensive experimental results show that SAT-LDM achieves robust watermarking while significantly improving the quality of watermarked images across diverse prompts. Furthermore, we conduct experimental analyses to demonstrate the strong generalization abilities of SAT-LDM. We hope our method offers a practical and convenient solution for securing high-fidelity AI-generated content.
Authors: Pedram Fekri, Mehrdad Zadeh, Javad Dargahi
Abstract: The success rate of catheterization procedures is closely linked to the sensory data provided to the surgeon. Vision-based deep learning models can deliver both tactile and visual information in a sensor-free manner, while also being cost-effective to produce. Given the complexity of these models for devices with limited computational resources, research has focused on force estimation and catheter segmentation separately. However, there is a lack of a comprehensive architecture capable of simultaneously segmenting the catheter from two different angles and estimating the applied forces in 3D. To bridge this gap, this work proposes a novel, lightweight, multi-input, multi-output encoder-decoder-based architecture. It is designed to segment the catheter from two points of view and concurrently measure the applied forces in the x, y, and z directions. This network processes two simultaneous X-Ray images, intended to be fed by a biplane fluoroscopy system, showing a catheter's deflection from different angles. It uses two parallel sub-networks with shared parameters to output two segmentation maps corresponding to the inputs. Additionally, it leverages stereo vision to estimate the applied forces at the catheter's tip in 3D. The architecture features two input channels, two classification heads for segmentation, and a regression head for force estimation through a single end-to-end architecture. The output of all heads was assessed and compared with the literature, demonstrating state-of-the-art performance in both segmentation and force estimation. To the best of the authors' knowledge, this is the first time such a model has been proposed
Authors: Asha V, Bhavanishankar K
Abstract: Lung cancer is an extremely lethal disease primarily due to its late-stage diagnosis and significant mortality rate, making it the major cause of cancer-related demises globally. Machine Learning (ML) and Convolution Neural network (CNN) based Deep Learning (DL) techniques are primarily used for precise segmentation and classification of cancerous nodules in the CT (Computed Tomography) or MRI images. This study introduces an innovative approach to lung nodule segmentation by utilizing the Segment Anything Model (SAM) combined with transfer learning techniques. Precise segmentation of lung nodules is crucial for the early detection of lung cancer. The proposed method leverages Bounding Box prompts and a vision transformer model to enhance segmentation performance, achieving high accuracy, Dice Similarity Coefficient (DSC) and Intersection over Union (IoU) metrics. The integration of SAM and Transfer Learning significantly improves Computer-Aided Detection (CAD) systems in medical imaging, particularly for lung cancer diagnosis. The findings demonstrate the proposed model effectiveness in precisely segmenting lung nodules from CT scans, underscoring its potential to advance early detection and improve patient care outcomes in lung cancer diagnosis. The results show SAM Model with transfer learning achieving a DSC of 97.08% and an IoU of 95.6%, for segmentation and accuracy of 96.71% for classification indicates that ,its performance is noteworthy compared to existing techniques.
Authors: Keith G. Mills, Muhammad Fetrat Qharabagh, Weichen Qiu, Fred X. Han, Mohammad Salameh, Wei Lu, Shangling Jui, Di Niu
Abstract: Layer fusion techniques are critical to improving the inference efficiency of deep neural networks (DNN) for deployment. Fusion aims to lower inference costs by reducing data transactions between an accelerator's on-chip buffer and DRAM. This is accomplished by grouped execution of multiple operations like convolution and activations together into single execution units - fusion groups. However, on-chip buffer capacity limits fusion group size and optimizing fusion on whole DNNs requires partitioning into multiple fusion groups. Finding the optimal groups is a complex problem where the presence of invalid solutions hampers traditional search algorithms and demands robust approaches. In this paper we incorporate Explainable AI, specifically Graph Explanation Techniques (GET), into layer fusion. Given an invalid fusion group, we identify the operations most responsible for group invalidity, then use this knowledge to recursively split the original fusion group via a greedy tree-based algorithm to minimize DRAM access. We pair our scheme with common algorithms and optimize DNNs on two types of layer fusion: Line-Buffer Depth First (LBDF) and Branch Requirement Reduction (BRR). Experiments demonstrate the efficacy of our scheme on several popular and classical convolutional neural networks like ResNets and MobileNets. Our scheme achieves over 20% DRAM Access reduction on EfficientNet-B3.
Authors: Abdesselam Ferdi
Abstract: Computer-aided diagnosis (CAD) systems have greatly improved the interpretation of medical images by radiologists and surgeons. However, current CAD systems for fracture detection in X-ray images primarily rely on large, resource-intensive detectors, which limits their practicality in clinical settings. To address this limitation, we propose a novel lightweight CAD system based on the YOLO detector for fracture detection. This system, named ghost convolution-based YOLOv11 (G-YOLOv11), builds on the latest version of the YOLO detector family and incorporates the ghost convolution operation for feature extraction. The ghost convolution operation generates the same number of feature maps as traditional convolution but requires fewer linear operations, thereby reducing the detector's computational resource requirements. We evaluated the performance of the proposed G-YOLOv11 detector on the GRAZPEDWRI-DX dataset, achieving an mAP@0.5 of 0.535 with an inference time of 2.4 ms on an NVIDIA A10 GPU. Compared to the standard YOLOv11l, G-YOLOv11l achieved reductions of 13.6% in mAP@0.5 and 68.7% in size. These results establish a new state-of-the-art benchmark in terms of efficiency, outperforming existing detectors. Code and models are available at https://github.com/AbdesselamFerdi/G-YOLOv11.
Authors: El Houcine El Fatimi
Abstract: This study, our main topic is to devlop a new deep-learning approachs for plant leaf disease identification and detection using leaf image datasets. We also discussed the challenges facing current methods of leaf disease detection and how deep learning may be used to overcome these challenges and enhance the accuracy of disease detection. Therefore, we have proposed a novel method for the detection of various leaf diseases in crops, along with the identification and description of an efficient network architecture that encompasses hyperparameters and optimization methods. The effectiveness of different architectures was compared and evaluated to see the best architecture configuration and to create an effective model that can quickly detect leaf disease. In addition to the work done on pre-trained models, we proposed a new model based on CNN, which provides an efficient method for identifying and detecting plant leaf disease. Furthermore, we evaluated the efficacy of our model and compared the results to those of some pre-trained state-of-the-art architectures.
Authors: HanQin Cai, Chandra Kundu, Jialin Liu, Wotao Yin
Abstract: Robust matrix completion (RMC) is a widely used machine learning tool that simultaneously tackles two critical issues in low-rank data analysis: missing data entries and extreme outliers. This paper proposes a novel scalable and learnable non-convex approach, coined Learned Robust Matrix Completion (LRMC), for large-scale RMC problems. LRMC enjoys low computational complexity with linear convergence. Motivated by the proposed theorem, the free parameters of LRMC can be effectively learned via deep unfolding to achieve optimum performance. Furthermore, this paper proposes a flexible feedforward-recurrent-mixed neural network framework that extends deep unfolding from fix-number iterations to infinite iterations. The superior empirical performance of LRMC is verified with extensive experiments against state-of-the-art on synthetic datasets and real applications, including video background subtraction, ultrasound imaging, face modeling, and cloud removal from satellite imagery.
Authors: Tetsuya Hoya, Shunpei Morita
Abstract: This paper proposes a novel approach to pattern classification using a probabilistic neural network model. The strategy is based on a compact-sized probabilistic neural network capable of continuous incremental learning and unlearning tasks. The network is constructed/reconstructed using a simple, one-pass network-growing algorithm with no hyperparameter tuning. Then, given the training dataset, its structure and parameters are automatically determined and can be dynamically varied in continual incremental and decremental learning situations. The algorithm proposed in this work involves no iterative or arduous matrix-based parameter approximations but a simple data-driven updating scheme. Simulation results using nine publicly available databases demonstrate the effectiveness of this approach, showing that compact-sized probabilistic neural networks constructed have a much smaller number of hidden units compared to the original probabilistic neural network model and yet can achieve a similar classification performance to that of multilayer perceptron neural networks in standard classification tasks, while also exhibiting sufficient capability in continuous class incremental learning and unlearning tasks.
Authors: Haoxuan Li, Wei song, Peiwu Qin, Xi Yuan, Zhenglin Chen
Abstract: Breast cancer lesion segmentation in DCE-MRI remains challenging due to heterogeneous tumor morphology and indistinct boundaries. To address these challenges, this study proposes a novel hybrid segmentation network, HCMA-UNet, for lesion segmentation of breast cancer. Our network consists of a lightweight CNN backbone and a Multi-view Inter-Slice Self-Attention Mamba (MISM) module. The MISM module integrates Visual State Space Block (VSSB) and Inter-Slice Self-Attention (ISSA) mechanism, effectively reducing parameters through Asymmetric Split Channel (ASC) strategy to achieve efficient tri-directional feature extraction. Our lightweight model achieves superior performance with 2.87M parameters and 126.44 GFLOPs. A Feature-guided Region-aware loss function (FRLoss) is proposed to enhance segmentation accuracy. Extensive experiments on one private and two public DCE-MRI breast cancer datasets demonstrate that our approach achieves state-of-the-art performance while maintaining computational efficiency. FRLoss also exhibits good cross-architecture generalization capabilities. The source code and dataset is available on this link.
Authors: Hirthik Mathesh GV, Kavin Chakravarthy M, Sentil Pandi S
Abstract: Oral cancer constitutes a significant global health concern, resulting in 277,484 fatalities in 2023, with the highest prevalence observed in low- and middle-income nations. Facilitating automation in the detection of possibly malignant and malignant lesions in the oral cavity could result in cost-effective and early disease diagnosis. Establishing an extensive repository of meticulously annotated oral lesions is essential. In this research photos are being collected from global clinical experts, who have been equipped with an annotation tool to generate comprehensive labelling. This research presents a novel approach for integrating bounding box annotations from various doctors. Additionally, Deep Belief Network combined with CAPSNET is employed to develop automated systems that extracted intricate patterns to address this challenging problem. This study evaluated two deep learning-based computer vision methodologies for the automated detection and classification of oral lesions to facilitate the early detection of oral cancer: image classification utilizing CAPSNET. Image classification attained an F1 score of 94.23% for detecting photos with lesions 93.46% for identifying images necessitating referral. Object detection attained an F1 score of 89.34% for identifying lesions for referral. Subsequent performances are documented about classification based on the sort of referral decision. Our preliminary findings indicate that deep learning possesses the capability to address this complex problem.
Authors: Junhuan Yang, Yuzhou Zhang, Yi Sheng, Youzuo Lin, Lei Yang
Abstract: Recently, the advent of generative AI technologies has made transformational impacts on our daily lives, yet its application in scientific applications remains in its early stages. Data scarcity is a major, well-known barrier in data-driven scientific computing, so physics-guided generative AI holds significant promise. In scientific computing, most tasks study the conversion of multiple data modalities to describe physical phenomena, for example, spatial and waveform in seismic imaging, time and frequency in signal processing, and temporal and spectral in climate modeling; as such, multi-modal pairwise data generation is highly required instead of single-modal data generation, which is usually used in natural images (e.g., faces, scenery). Moreover, in real-world applications, the unbalance of available data in terms of modalities commonly exists; for example, the spatial data (i.e., velocity maps) in seismic imaging can be easily simulated, but real-world seismic waveform is largely lacking. While the most recent efforts enable the powerful diffusion model to generate multi-modal data, how to leverage the unbalanced available data is still unclear. In this work, we use seismic imaging in subsurface geophysics as a vehicle to present ``UB-Diff'', a novel diffusion model for multi-modal paired scientific data generation. One major innovation is a one-in-two-out encoder-decoder network structure, which can ensure pairwise data is obtained from a co-latent representation. Then, the co-latent representation will be used by the diffusion process for pairwise data generation. Experimental results on the OpenFWI dataset show that UB-Diff significantly outperforms existing techniques in terms of Fr\'{e}chet Inception Distance (FID) score and pairwise evaluation, indicating the generation of reliable and useful multi-modal pairwise data.
Authors: Lukas Kuhn, Sari Sadiya, Jorg Schlotterer, Christin Seifert, Gemma Roig
Abstract: Shortcut learning, i.e., a model's reliance on undesired features not directly relevant to the task, is a major challenge that severely limits the applications of machine learning algorithms, particularly when deploying them to assist in making sensitive decisions, such as in medical diagnostics. In this work, we leverage recent advancements in machine learning to create an unsupervised framework that is capable of both detecting and mitigating shortcut learning in transformers. We validate our method on multiple datasets. Results demonstrate that our framework significantly improves both worst-group accuracy (samples misclassified due to shortcuts) and average accuracy, while minimizing human annotation effort. Moreover, we demonstrate that the detected shortcuts are meaningful and informative to human experts, and that our framework is computationally efficient, allowing it to be run on consumer hardware.
Authors: Sagarnil Das, Pradeep Walia
Abstract: Diabetic Retinopathy (DR) is a leading cause of preventable blindness. Early detection at the DR1 stage is critical but is hindered by a scarcity of high-quality fundus images. This study uses StyleGAN3 to generate synthetic DR1 images characterized by microaneurysms with high fidelity and diversity. The aim is to address data scarcity and enhance the performance of supervised classifiers. A dataset of 2,602 DR1 images was used to train the model, followed by a comprehensive evaluation using quantitative metrics, including Frechet Inception Distance (FID), Kernel Inception Distance (KID), and Equivariance with respect to translation (EQ-T) and rotation (EQ-R). Qualitative assessments included Human Turing tests, where trained ophthalmologists evaluated the realism of synthetic images. Spectral analysis further validated image quality. The model achieved a final FID score of 17.29, outperforming the mean FID of 21.18 (95 percent confidence interval - 20.83 to 21.56) derived from bootstrap resampling. Human Turing tests demonstrated the model's ability to produce highly realistic images, though minor artifacts near the borders were noted. These findings suggest that StyleGAN3-generated synthetic DR1 images hold significant promise for augmenting training datasets, enabling more accurate early detection of Diabetic Retinopathy. This methodology highlights the potential of synthetic data in advancing medical imaging and AI-driven diagnostics.
Authors: Chenyu You, Haocheng Dai, Yifei Min, Jasjeet S. Sekhon, Sarang Joshi, James S. Duncan
Abstract: Machine learning models often rely on simple spurious features -- patterns in training data that correlate with targets but are not causally related to them, like image backgrounds in foreground classification. This reliance typically leads to imbalanced test performance across minority and majority groups. In this work, we take a closer look at the fundamental cause of such imbalanced performance through the lens of memorization, which refers to the ability to predict accurately on \textit{atypical} examples (minority groups) in the training set but failing in achieving the same accuracy in the testing set. This paper systematically shows the ubiquitous existence of spurious features in a small set of neurons within the network, providing the first-ever evidence that memorization may contribute to imbalanced group performance. Through three experimental sources of converging empirical evidence, we find the property of a small subset of neurons or channels in memorizing minority group information. Inspired by these findings, we articulate the hypothesis: the imbalanced group performance is a byproduct of ``noisy'' spurious memorization confined to a small set of neurons. To further substantiate this hypothesis, we show that eliminating these unnecessary spurious memorization patterns via a novel framework during training can significantly affect the model performance on minority groups. Our experimental results across various architectures and benchmarks offer new insights on how neural networks encode core and spurious knowledge, laying the groundwork for future research in demystifying robustness to spurious correlation.
Authors: Xiaoshuai Hao, Guanqun Liu, Yuting Zhao, Yuheng Ji, Mengchuan Wei, Haimei Zhao, Lingdong Kong, Rong Yin, Yu Liu
Abstract: Multi-sensor fusion models play a crucial role in autonomous driving perception, particularly in tasks like 3D object detection and HD map construction. These models provide essential and comprehensive static environmental information for autonomous driving systems. While camera-LiDAR fusion methods have shown promising results by integrating data from both modalities, they often depend on complete sensor inputs. This reliance can lead to low robustness and potential failures when sensors are corrupted or missing, raising significant safety concerns. To tackle this challenge, we introduce the Multi-Sensor Corruption Benchmark (MSC-Bench), the first comprehensive benchmark aimed at evaluating the robustness of multi-sensor autonomous driving perception models against various sensor corruptions. Our benchmark includes 16 combinations of corruption types that disrupt both camera and LiDAR inputs, either individually or concurrently. Extensive evaluations of six 3D object detection models and four HD map construction models reveal substantial performance degradation under adverse weather conditions and sensor failures, underscoring critical safety issues. The benchmark toolkit and affiliated code and model checkpoints have been made publicly accessible.
Authors: Syed Tahir Hussain Rizvi, Neel Kanwal, Muddasar Naeem, Alfredo Cuzzocrea, Antonio Coronato
Abstract: Time Series Forecasting (TSF) is an important application across many fields. There is a debate about whether Transformers, despite being good at understanding long sequences, struggle with preserving temporal relationships in time series data. Recent research suggests that simpler linear models might outperform or at least provide competitive performance compared to complex Transformer-based models for TSF tasks. In this paper, we propose a novel data-efficient architecture, GLinear, for multivariate TSF that exploits periodic patterns to provide better accuracy. It also provides better prediction accuracy by using a smaller amount of historical data compared to other state-of-the-art linear predictors. Four different datasets (ETTh1, Electricity, Traffic, and Weather) are used to evaluate the performance of the proposed predictor. A performance comparison with state-of-the-art linear architectures (such as NLinear, DLinear, and RLinear) and transformer-based time series predictor (Autoformer) shows that the GLinear, despite being parametrically efficient, significantly outperforms the existing architectures in most cases of multivariate TSF. We hope that the proposed GLinear opens new fronts of research and development of simpler and more sophisticated architectures for data and computationally efficient time-series analysis. The source code is publicly available on GitHub.
Authors: Yixu Wang, Tianle Gu, Yan Teng, Yingchun Wang, Xingjun Ma
Abstract: Model extraction attacks are one type of inference-time attacks that approximate the functionality and performance of a black-box victim model by launching a certain number of queries to the model and then leveraging the model's predictions to train a substitute model. These attacks pose severe security threats to production models and MLaaS platforms and could cause significant monetary losses to the model owners. A body of work has proposed to defend machine learning models against model extraction attacks, including both active defense methods that modify the model's outputs or increase the query overhead to avoid extraction and passive defense methods that detect malicious queries or leverage watermarks to perform post-verification. In this work, we introduce a new defense paradigm called attack as defense which modifies the model's output to be poisonous such that any malicious users that attempt to use the output to train a substitute model will be poisoned. To this end, we propose a novel lightweight backdoor attack method dubbed HoneypotNet that replaces the classification layer of the victim model with a honeypot layer and then fine-tunes the honeypot layer with a shadow model (to simulate model extraction) via bi-level optimization to modify its output to be poisonous while remaining the original performance. We empirically demonstrate on four commonly used benchmark datasets that HoneypotNet can inject backdoors into substitute models with a high success rate. The injected backdoor not only facilitates ownership verification but also disrupts the functionality of substitute models, serving as a significant deterrent to model extraction attacks.
Authors: Francisco Mena, Diego Arenas, Andreas Dengel
Abstract: Multi-view learning (MVL) leverages multiple sources or views of data to enhance machine learning model performance and robustness. This approach has been successfully used in the Earth Observation (EO) domain, where views have a heterogeneous nature and can be affected by missing data. Despite the negative effect that missing data has on model predictions, the ML literature has used it as an augmentation technique to improve model generalization, like masking the input data. Inspired by this, we introduce novel methods for EO applications tailored to MVL with missing views. Our methods integrate the combination of a set to simulate all combinations of missing views as different training samples. Instead of replacing missing data with a numerical value, we use dynamic merge functions, like average, and more complex ones like Transformer. This allows the MVL model to entirely ignore the missing views, enhancing its predictive robustness. We experiment on four EO datasets with temporal and static views, including state-of-the-art methods from the EO domain. The results indicate that our methods improve model robustness under conditions of moderate missingness, and improve the predictive performance when all views are present. The proposed methods offer a single adaptive solution to operate effectively with any combination of available views.
Authors: Shudong Liu, Yiqiao Jin, Cheng Li, Derek F. Wong, Qingsong Wen, Lichao Sun, Haipeng Chen, Xing Xie, Jindong Wang
Abstract: Vision-language models (VLMs) have advanced human-AI interaction but struggle with cultural understanding, often misinterpreting symbols, gestures, and artifacts due to biases in predominantly Western-centric training data. In this paper, we construct CultureVerse, a large-scale multimodal benchmark covering 19, 682 cultural concepts, 188 countries/regions, 15 cultural concepts, and 3 question types, with the aim of characterizing and improving VLMs' multicultural understanding capabilities. Then, we propose CultureVLM, a series of VLMs fine-tuned on our dataset to achieve significant performance improvement in cultural understanding. Our evaluation of 16 models reveals significant disparities, with a stronger performance in Western concepts and weaker results in African and Asian contexts. Fine-tuning on our CultureVerse enhances cultural perception, demonstrating cross-cultural, cross-continent, and cross-dataset generalization without sacrificing performance on models' general VLM benchmarks. We further present insights on cultural generalization and forgetting. We hope that this work could lay the foundation for more equitable and culturally aware multimodal AI systems.
Authors: Shuya Tobari, Shuhei Tomoshige, Hayato Muraki, Kenichi Oishi, Hitoshi Iyatomi
Abstract: When conducting large-scale studies that collect brain MR images from multiple facilities, the impact of differences in imaging equipment and protocols at each site cannot be ignored, and this domain gap has become a significant issue in recent years. In this study, we propose a new low-dimensional representation (LDR) acquisition method called style encoder adversarial domain adaptation (SE-ADA) to realize content-based image retrieval (CBIR) of brain MR images. SE-ADA reduces domain differences while preserving pathological features by separating domain-specific information from LDR and minimizing domain differences using adversarial learning. In evaluation experiments comparing SE-ADA with recent domain harmonization methods on eight public brain MR datasets (ADNI1/2/3, OASIS1/2/3/4, PPMI), SE-ADA effectively removed domain information while preserving key aspects of the original brain structure and demonstrated the highest disease search accuracy.
Authors: Neda Tavakoli, Amir Ali Rahsepar, Brandon C. Benefield, Daming Shen, Santiago L\'opez-Tapia, Florian Schiffers, Jeffrey J. Goldberger, Christine M. Albert, Edwin Wu, Aggelos K. Katsaggelos, Daniel C. Lee, Daniel Kim
Abstract: Background: Late Gadolinium Enhancement (LGE) imaging is the gold standard for assessing myocardial fibrosis and scarring, with left ventricular (LV) LGE extent predicting major adverse cardiac events (MACE). Despite its importance, routine LGE-based LV scar quantification is hindered by labor-intensive manual segmentation and inter-observer variability. Methods: We propose ScarNet, a hybrid model combining a transformer-based encoder from the Medical Segment Anything Model (MedSAM) with a convolution-based U-Net decoder, enhanced by tailored attention blocks. ScarNet was trained on 552 ischemic cardiomyopathy patients with expert segmentations of myocardial and scar boundaries and tested on 184 separate patients. Results: ScarNet achieved robust scar segmentation in 184 test patients, yielding a median Dice score of 0.912 (IQR: 0.863--0.944), significantly outperforming MedSAM (median Dice = 0.046, IQR: 0.043--0.047) and nnU-Net (median Dice = 0.638, IQR: 0.604--0.661). ScarNet demonstrated lower bias (-0.63%) and coefficient of variation (4.3%) compared to MedSAM (bias: -13.31%, CoV: 130.3%) and nnU-Net (bias: -2.46%, CoV: 20.3%). In Monte Carlo simulations with noise perturbations, ScarNet achieved significantly higher scar Dice (0.892 \pm 0.053, CoV = 5.9%) than MedSAM (0.048 \pm 0.112, CoV = 233.3%) and nnU-Net (0.615 \pm 0.537, CoV = 28.7%). Conclusion: ScarNet outperformed MedSAM and nnU-Net in accurately segmenting myocardial and scar boundaries in LGE images. The model exhibited robust performance across diverse image qualities and scar patterns.
Authors: Yucheng Zhou, Lingran Song, Jianbing Shen
Abstract: Existing Medical Large Vision-Language Models (Med-LVLMs), which encapsulate extensive medical knowledge, demonstrate excellent capabilities in understanding medical images and responding to human queries based on these images. However, there remain challenges in visual localization in medical images, which is crucial for abnormality detection and interpretation. To address these issues, we propose a novel UMed-LVLM designed with Unveiling Medical abnormalities. Specifically, we collect a Medical Abnormalities Unveiling (MAU) dataset and propose a two-stage training method for UMed-LVLM training. To collect MAU dataset, we propose a prompt method utilizing the GPT-4V to generate diagnoses based on identified abnormal areas in medical images. Moreover, the two-stage training method includes Abnormal-Aware Instruction Tuning and Abnormal-Aware Rewarding, comprising Abnormal Localization Rewarding and Vision Relevance Rewarding. Experimental results demonstrate that our UMed-LVLM surpasses existing Med-LVLMs in identifying and understanding medical abnormality. In addition, this work shows that enhancing the abnormality detection capabilities of Med-LVLMs significantly improves their understanding of medical images and generalization capability.
Authors: Xuyin Qi, Zeyu Zhang, Aaron Berliano Handoko, Huazhan Zheng, Mingxi Chen, Ta Duc Huy, Vu Minh Hieu Phan, Lei Zhang, Linqi Cheng, Shiyu Jiang, Zhiwei Zhang, Zhibin Liao, Yang Zhao, Minh-Son To
Abstract: Prostate cancer, a growing global health concern, necessitates precise diagnostic tools, with Magnetic Resonance Imaging (MRI) offering high-resolution soft tissue imaging that significantly enhances diagnostic accuracy. Recent advancements in explainable AI and representation learning have significantly improved prostate cancer diagnosis by enabling automated and precise lesion classification. However, existing explainable AI methods, particularly those based on frameworks like generative adversarial networks (GANs), are predominantly developed for natural image generation, and their application to medical imaging often leads to suboptimal performance due to the unique characteristics and complexity of medical image. To address these challenges, our paper introduces three key contributions. First, we propose ProjectedEx, a generative framework that provides interpretable, multi-attribute explanations, effectively linking medical image features to classifier decisions. Second, we enhance the encoder module by incorporating feature pyramids, which enables multiscale feedback to refine the latent space and improves the quality of generated explanations. Additionally, we conduct comprehensive experiments on both the generator and classifier, demonstrating the clinical relevance and effectiveness of ProjectedEx in enhancing interpretability and supporting the adoption of AI in medical settings. Code will be released at https://github.com/Richardqiyi/ProjectedEx
Authors: Yuanpeng Tu, Yuxi Li, Boshen Zhang, Liang Liu, Jiangning Zhang, Yabiao Wang, Cai Rong Zhao
Abstract: Robust autonomous driving requires agents to accurately identify unexpected areas (anomalies) in urban scenes. To this end, some critical issues remain open: how to design advisable metric to measure anomalies, and how to properly generate training samples of anomaly data? Classical effort in anomaly detection usually resorts to pixel-wise uncertainty or sample synthesis, which ignores the contextual information and sometimes requires auxiliary data with fine-grained annotations. On the contrary, in this paper, we exploit the strong context-dependent nature of the segmentation task and design an energy-guided self-supervised framework for anomaly segmentation, which optimizes an anomaly head by maximizing the likelihood of self-generated anomaly pixels. For this purpose, we design two estimators to model anomaly likelihood, one is a task-agnostic binary estimator and the other depicts the likelihood as residual of task-oriented joint energy. Based on the proposed estimators, we devise an adaptive self-supervised training framework, which exploits the contextual reliance and estimated likelihood to refine mask annotations in anomaly areas. We conduct extensive experiments on challenging Fishyscapes and Road Anomaly benchmarks, demonstrating that without any auxiliary data or synthetic models, our method can still achieve comparable performance to supervised competitors. Code is available at https://github.com/yuanpengtu/SLEEG..
Authors: Weihao Yu, Pan Zhou, Shuicheng Yan, Xinchao Wang
Abstract: Inspired by the long-range modeling ability of ViTs, large-kernel convolutions are widely studied and adopted recently to enlarge the receptive field and improve model performance, like the remarkable work ConvNeXt which employs 7x7 depthwise convolution. Although such depthwise operator only consumes a few FLOPs, it largely harms the model efficiency on powerful computing devices due to the high memory access costs. For example, ConvNeXt-T has similar FLOPs with ResNet-50 but only achieves ~60% throughputs when trained on A100 GPUs with full precision. Although reducing the kernel size of ConvNeXt can improve speed, it results in significant performance degradation, which poses a challenging problem: How to speed up large-kernel-based CNN models while preserving their performance. To tackle this issue, inspired by Inceptions, we propose to decompose large-kernel depthwise convolution into four parallel branches along channel dimension, i.e., small square kernel, two orthogonal band kernels, and an identity mapping. With this new Inception depthwise convolution, we build a series of networks, namely IncepitonNeXt, which not only enjoy high throughputs but also maintain competitive performance. For instance, InceptionNeXt-T achieves 1.6x higher training throughputs than ConvNeX-T, as well as attains 0.2% top-1 accuracy improvement on ImageNet-1K. We anticipate InceptionNeXt can serve as an economical baseline for future architecture design to reduce carbon footprint. Code is available at https://github.com/sail-sg/inceptionnext.
Authors: Qi Wu, Mingyan Han, Ting Jiang, Chengzhi Jiang, Jinting Luo, Man Jiang, Haoqiang Fan, Shuaicheng Liu
Abstract: Deep denoising models require extensive real-world training data, which is challenging to acquire. Current noise synthesis techniques struggle to accurately model complex noise distributions. We propose a novel Realistic Noise Synthesis Diffusor (RNSD) method using diffusion models to address these challenges. By encoding camera settings into a time-aware camera-conditioned affine modulation (TCCAM), RNSD generates more realistic noise distributions under various camera conditions. Additionally, RNSD integrates a multi-scale content-aware module (MCAM), enabling the generation of structured noise with spatial correlations across multiple frequencies. We also introduce Deep Image Prior Sampling (DIPS), a learnable sampling sequence based on depth image prior, which significantly accelerates the sampling process while maintaining the high quality of synthesized noise. Extensive experiments demonstrate that our RNSD method significantly outperforms existing techniques in synthesizing realistic noise under multiple metrics and improving image denoising performance.
Authors: Huake Wang, Xingsong Hou, Chengcu Liu, Kaibing Zhang, Xiangyong Cao, Xueming Qian
Abstract: Although low-light image enhancement has achieved great stride based on deep enhancement models, most of them mainly stress on enhancement performance via an elaborated black-box network and rarely explore the physical significance of enhancement models. Towards this issue, we propose a Dual degrAdation-inSpired deep Unfolding network, termed DASUNet, for low-light image enhancement. Specifically, we construct a dual degradation model (DDM) to explicitly simulate the deterioration mechanism of low-light images. It learns two distinct image priors via considering degradation specificity between luminance and chrominance spaces. To make the proposed scheme tractable, we design an alternating optimization solution to solve the proposed DDM. Further, the designed solution is unfolded into a specified deep network, imitating the iteration updating rules, to form DASUNet. Based on different specificity in two spaces, we design two customized Transformer block to model different priors. Additionally, a space aggregation module (SAM) is presented to boost the interaction of two degradation models. Extensive experiments on multiple popular low-light image datasets validate the effectiveness of DASUNet compared to canonical state-of-the-art low-light image enhancement methods. Our source code and pretrained model will be publicly available.
Authors: Xiaoyan Cao, Yiyao Zheng, Yao Yao, Huapeng Qin, Xiaoyu Cao, Shihui Guo
Abstract: Video data and algorithms have been driving advances in multi-object tracking (MOT). While existing MOT datasets focus on occlusion and appearance similarity, complex motion patterns are widespread yet overlooked. To address this issue, we introduce a new dataset called BEE24 to highlight complex motions. Identity association algorithms have long been the focus of MOT research. Existing trackers can be categorized into two association paradigms: single-feature paradigm (based on either motion or appearance feature) and serial paradigm (one feature serves as secondary while the other is primary). However, these paradigms are incapable of fully utilizing different features. In this paper, we propose a parallel paradigm and present the Two rOund Parallel matchIng meChanism (TOPIC) to implement it. The TOPIC leverages both motion and appearance features and can adaptively select the preferable one as the assignment metric based on motion level. Moreover, we provide an Attention-based Appearance Reconstruction Module (AARM) to reconstruct appearance feature embeddings, thus enhancing the representation of appearance features. Comprehensive experiments show that our approach achieves state-of-the-art performance on four public datasets and BEE24. Moreover, BEE24 challenges existing trackers to track multiple similar-appearing small objects with complex motions over long periods, which is critical in real-world applications such as beekeeping and drone swarm surveillance. Notably, our proposed parallel paradigm surpasses the performance of existing association paradigms by a large margin, e.g., reducing false negatives by 6% to 81% compared to the single-feature association paradigm. The introduced dataset and association paradigm in this work offer a fresh perspective for advancing the MOT field. The source code and dataset are available at https://github.com/holmescao/TOPICTrack.
Authors: Philip Enevoldsen, Christian Gundersen, Nico Lang, Serge Belongie, Christian Igel
Abstract: Open-set recognition (OSR), the identification of novel categories, can be a critical component when deploying classification models in real-world applications. Recent work has shown that familiarity-based scoring rules such as the Maximum Softmax Probability (MSP) or the Maximum Logit Score (MLS) are strong baselines when the closed-set accuracy is high. However, one of the potential weaknesses of familiarity-based OSR are adversarial attacks. Here, we study gradient-based adversarial attacks on familiarity scores for both types of attacks, False Familiarity and False Novelty attacks, and evaluate their effectiveness in informed and uninformed settings on TinyImageNet. Furthermore, we explore how novel and familiar samples react to adversarial attacks and formulate the adversarial reaction score as an alternative OSR scoring rule, which shows a high correlation with the MLS familiarity score.
Authors: Chi-Ping Su, Ching-Hsun Tseng, Bin Pu, Lei Zhao, Zhuangzhuang Chen, Shin-Jye Lee
Abstract: Knowledge distillation (KD) enables a smaller "student" model to mimic a larger "teacher" model by transferring knowledge from the teacher's output or features. However, most KD methods treat all samples uniformly, overlooking the varying learning value of each sample and thereby limiting effectiveness. In this paper, we propose Entropy-based Adaptive Knowledge Distillation (EA-KD), a simple yet effective plug-and-play KD method that prioritizes learning from valuable samples. EA-KD quantifies each sample's learning value by strategically combining the entropy of the teacher and student output, then dynamically reweights the distillation loss to place greater emphasis on high-value samples. Extensive experiments across diverse KD frameworks and tasks$\unicode{x2014}$including image classification, object detection, and large language model (LLM) distillation$\unicode{x2014}$demonstrate that EA-KD consistently enhances performance, achieving state-of-the-art results with negligible computational cost. Our code will be publicly available.
Authors: Viktor Kocur, Daniel Kyselica, Zuzana K\'ukelov\'a
Abstract: The problem of self-calibration of two cameras from a given fundamental matrix is one of the basic problems in geometric computer vision. Under the assumption of known principal points and square pixels, the well-known Bougnoux formula offers a means to compute the two unknown focal lengths. However, in many practical situations, the formula yields inaccurate results due to commonly occurring singularities. Moreover, the estimates are sensitive to noise in the computed fundamental matrix and to the assumed positions of the principal points. In this paper, we therefore propose an efficient and robust iterative method to estimate the focal lengths along with the principal points of the cameras given a fundamental matrix and priors for the estimated camera parameters. In addition, we study a computationally efficient check of models generated within RANSAC that improves the accuracy of the estimated models while reducing the total computational time. Extensive experiments on real and synthetic data show that our iterative method brings significant improvements in terms of the accuracy of the estimated focal lengths over the Bougnoux formula and other state-of-the-art methods, even when relying on inaccurate priors.
Authors: Zihao Wang, Jing Zhao, Xuetong Ding, Hui Zhang
Abstract: Face plays an important role in humans visual perception, and reconstructing perceived faces from brain activities is challenging because of its difficulty in extracting high-level features and maintaining consistency of multiple face attributes, such as expression, identity, gender, etc. In this study, we proposed a novel reconstruction framework, which we called Double-Flow GAN, that can enhance the capability of discriminator and handle imbalances in images from certain domains that are too easy for generators. We also designed a pretraining process that uses features extracted from images as conditions for making it possible to pretrain the conditional reconstruction model from fMRI in a larger pure image dataset. Moreover, we developed a simple pretrained model for fMRI alignment to alleviate the problem of cross-subject reconstruction due to the variations of brain structure among different subjects. We conducted experiments by using our proposed method and traditional reconstruction models. Results showed that the proposed method is significant at accurately reconstructing multiple face attributes, outperforms the previous reconstruction models, and exhibited state-of-the-art reconstruction abilities.
Authors: Jiarui Zhang, Ruixu Geng, Xiaolong Du, Yan Chen, Houqiang Li, Yang Hu
Abstract: Passive non-line-of-sight (NLOS) imaging has witnessed rapid development in recent years, due to its ability to image objects that are out of sight. The light transport condition plays an important role in this task since changing the conditions will lead to different imaging models. Existing learning-based NLOS methods usually train independent models for different light transport conditions, which is computationally inefficient and impairs the practicality of the models. In this work, we propose NLOS-LTM, a novel passive NLOS imaging method that effectively handles multiple light transport conditions with a single network. We achieve this by inferring a latent light transport representation from the projection image and using this representation to modulate the network that reconstructs the hidden image from the projection image. We train a light transport encoder together with a vector quantizer to obtain the light transport representation. To further regulate this representation, we jointly learn both the reconstruction network and the reprojection network during training. A set of light transport modulation blocks is used to modulate the two jointly trained networks in a multi-scale way. Extensive experiments on a large-scale passive NLOS dataset demonstrate the superiority of the proposed method. The code is available at https://github.com/JerryOctopus/NLOS-LTM.
Authors: Chih-Hao Lin, Jia-Bin Huang, Zhengqin Li, Zhao Dong, Christian Richardt, Tuotuo Li, Michael Zollh\"ofer, Johannes Kopf, Shenlong Wang, Changil Kim
Abstract: Inverse rendering seeks to recover 3D geometry, surface material, and lighting from captured images, enabling advanced applications such as novel-view synthesis, relighting, and virtual object insertion. However, most existing techniques rely on high dynamic range (HDR) images as input, limiting accessibility for general users. In response, we introduce IRIS, an inverse rendering framework that recovers the physically based material, spatially-varying HDR lighting, and camera response functions from multi-view, low-dynamic-range (LDR) images. By eliminating the dependence on HDR input, we make inverse rendering technology more accessible. We evaluate our approach on real-world and synthetic scenes and compare it with state-of-the-art methods. Our results show that IRIS effectively recovers HDR lighting, accurate material, and plausible camera response functions, supporting photorealistic relighting and object insertion.
Authors: Xinyue Xu, Yi Qin, Lu Mi, Hao Wang, Xiaomeng Li
Abstract: Existing methods, such as concept bottleneck models (CBMs), have been successful in providing concept-based interpretations for black-box deep learning models. They typically work by predicting concepts given the input and then predicting the final class label given the predicted concepts. However, (1) they often fail to capture the high-order, nonlinear interaction between concepts, e.g., correcting a predicted concept (e.g., "yellow breast") does not help correct highly correlated concepts (e.g., "yellow belly"), leading to suboptimal final accuracy; (2) they cannot naturally quantify the complex conditional dependencies between different concepts and class labels (e.g., for an image with the class label "Kentucky Warbler" and a concept "black bill", what is the probability that the model correctly predicts another concept "black crown"), therefore failing to provide deeper insight into how a black-box model works. In response to these limitations, we propose Energy-based Concept Bottleneck Models (ECBMs). Our ECBMs use a set of neural networks to define the joint energy of candidate (input, concept, class) tuples. With such a unified interface, prediction, concept correction, and conditional dependency quantification are then represented as conditional probabilities, which are generated by composing different energy functions. Our ECBMs address both limitations of existing CBMs, providing higher accuracy and richer concept interpretations. Empirical results show that our approach outperforms the state-of-the-art on real-world datasets.
Authors: Yongkun Du, Zhineng Chen, Yuchen Su, Caiyan Jia, Yu-Gang Jiang
Abstract: Multi-modal models have shown appealing performance in visual recognition tasks, as free-form text-guided training evokes the ability to understand fine-grained visual content. However, current models cannot be trivially applied to scene text recognition (STR) due to the compositional difference between natural and text images. We propose a novel instruction-guided scene text recognition (IGTR) paradigm that formulates STR as an instruction learning problem and understands text images by predicting character attributes, e.g., character frequency, position, etc. IGTR first devises $\left \langle condition,question,answer\right \rangle$ instruction triplets, providing rich and diverse descriptions of character attributes. To effectively learn these attributes through question-answering, IGTR develops a lightweight instruction encoder, a cross-modal feature fusion module and a multi-task answer head, which guides nuanced text image understanding. Furthermore, IGTR realizes different recognition pipelines simply by using different instructions, enabling a character-understanding-based text reasoning paradigm that differs from current methods considerably. Experiments on English and Chinese benchmarks show that IGTR outperforms existing models by significant margins, while maintaining a small model size and fast inference speed. Moreover, by adjusting the sampling of instructions, IGTR offers an elegant way to tackle the recognition of rarely appearing and morphologically similar characters, which were previous challenges. Code: https://github.com/Topdu/OpenOCR.
Authors: Ruyi Lian, Yuewei Lin, Longin Jan Latecki, Haibin Ling
Abstract: Localizing predefined 3D keypoints in a 2D image is an effective way to establish 3D-2D correspondences for 6DoF object pose estimation. However, unreliable localization results of invisible keypoints degrade the quality of correspondences. In this paper, we address this issue by localizing the important keypoints in terms of visibility. Since keypoint visibility information is currently missing in the dataset collection process, we propose an efficient way to generate binary visibility labels from available object-level annotations, for keypoints of both asymmetric objects and symmetric objects. We further derive real-valued visibility-aware importance from binary labels based on the PageRank algorithm. Taking advantage of the flexibility of our visibility-aware importance, we construct VAPO (Visibility-Aware POse estimator) by integrating the visibility-aware importance with a state-of-the-art pose estimation algorithm, along with additional positional encoding. VAPO can work in both CAD-based and CAD-free settings. Extensive experiments are conducted on popular pose estimation benchmarks including Linemod, Linemod-Occlusion, and YCB-V. The results show that, VAPO significantly improves both the keypoint correspondences and final estimated poses, and clearly achieves state-of-the-art performances.
Authors: Brian Hsuan-Cheng Liao, Chih-Hong Cheng, Hasan Esen, Alois Knoll
Abstract: This paper presents Ego-Centric Intersection-over-Union (EC-IoU), addressing the limitation of the standard IoU measure in characterizing safety-related performance for object detectors in navigating contexts. Concretely, we propose a weighting mechanism to refine IoU, allowing it to assign a higher score to a prediction that covers closer points of a ground-truth object from the ego agent's perspective. The proposed EC-IoU measure can be used in typical evaluation processes to select object detectors with better safety-related performance for downstream tasks. It can also be integrated into common loss functions for model fine-tuning. While geared towards safety, our experiment with the KITTI dataset demonstrates the performance of a model trained on EC-IoU can be better than that of a variant trained on IoU in terms of mean Average Precision as well.
Authors: Sibeak Lee, Kyeongsu Kang, Seongbo Ha, Hyeonwoo Yu
Abstract: We present a Bayesian Neural Radiance Field (NeRF), which explicitly quantifies uncertainty in the volume density by modeling uncertainty in the occupancy, without the need for additional networks, making it particularly suited for challenging observations and uncontrolled image environments. NeRF diverges from traditional geometric methods by providing an enriched scene representation, rendering color and density in 3D space from various viewpoints. However, NeRF encounters limitations in addressing uncertainties solely through geometric structure information, leading to inaccuracies when interpreting scenes with insufficient real-world observations. While previous efforts have relied on auxiliary networks, we propose a series of formulation extensions to NeRF that manage uncertainties in density, both color and density, and occupancy, all without the need for additional networks. In experiments, we show that our method significantly enhances performance on RGB and depth images in the comprehensive dataset. Given that uncertainty modeling aligns well with the inherently uncertain environments of Simultaneous Localization and Mapping (SLAM), we applied our approach to SLAM systems and observed notable improvements in mapping and tracking performance. These results confirm the effectiveness of our Bayesian NeRF approach in quantifying uncertainty based on geometric structure, making it a robust solution for challenging real-world scenarios.
Authors: Yuzhu Cai, Sheng Yin, Yuxi Wei, Chenxin Xu, Weibo Mao, Felix Juefei-Xu, Siheng Chen, Yanfeng Wang
Abstract: The burgeoning landscape of text-to-image models, exemplified by innovations such as Midjourney and DALLE 3, has revolutionized content creation across diverse sectors. However, these advancements bring forth critical ethical concerns, particularly with the misuse of open-source models to generate content that violates societal norms. Addressing this, we introduce Ethical-Lens, a framework designed to facilitate the value-aligned usage of text-to-image tools without necessitating internal model revision. Ethical-Lens ensures value alignment in text-to-image models across toxicity and bias dimensions by refining user commands and rectifying model outputs. Systematic evaluation metrics, combining GPT4-V, HEIM, and FairFace scores, assess alignment capability. Our experiments reveal that Ethical-Lens enhances alignment capabilities to levels comparable with or superior to commercial models like DALLE 3, ensuring user-generated content adheres to ethical standards while maintaining image quality. This study indicates the potential of Ethical-Lens to ensure the sustainable development of open-source text-to-image tools and their beneficial integration into society. Our code is available at https://github.com/yuzhu-cai/Ethical-Lens.
Authors: Rahul Sajnani, Jeroen Vanbaar, Jie Min, Kapil Katyal, Srinath Sridhar
Abstract: The success of image generative models has enabled us to build methods that can edit images based on text or other user input. However, these methods are bespoke, imprecise, require additional information, or are limited to only 2D image edits. We present GeoDiffuser, a zero-shot optimization-based method that unifies common 2D and 3D image-based object editing capabilities into a single method. Our key insight is to view image editing operations as geometric transformations. We show that these transformations can be directly incorporated into the attention layers in diffusion models to implicitly perform editing operations. Our training-free optimization method uses an objective function that seeks to preserve object style but generate plausible images, for instance with accurate lighting and shadows. It also inpaints disoccluded parts of the image where the object was originally located. Given a natural image and user input, we segment the foreground object using SAM and estimate a corresponding transform which is used by our optimization approach for editing. GeoDiffuser can perform common 2D and 3D edits like object translation, 3D rotation, and removal. We present quantitative results, including a perceptual study, that shows how our approach is better than existing methods. Visit https://ivl.cs.brown.edu/research/geodiffuser.html for more information.
Authors: Yiming Wu, Qihe Pan, Zhen Zhao, Zicheng Wang, Sifan Long, Ronghua Liang
Abstract: In this paper, we delve into a new task known as small object editing (SOE), which focuses on text-based image inpainting within a constrained, small-sized area. Despite the remarkable success have been achieved by current image inpainting approaches, their application to the SOE task generally results in failure cases such as Object Missing, Text-Image Mismatch, and Distortion. These failures stem from the limited use of small-sized objects in training datasets and the downsampling operations employed by U-Net models, which hinders accurate generation. To overcome these challenges, we introduce a novel training-based approach, SOEDiff, aimed at enhancing the capability of baseline models like StableDiffusion in editing small-sized objects while minimizing training costs. Specifically, our method involves two key components: SO-LoRA, which efficiently fine-tunes low-rank matrices, and Cross-Scale Score Distillation loss, which leverages high-resolution predictions from the pre-trained teacher diffusion model. Our method presents significant improvements on the test dataset collected from MSCOCO and OpenImage, validating the effectiveness of our proposed method in small object editing. In particular, when comparing SOEDiff with SD-I model on the OpenImage-f dataset, we observe a 0.99 improvement in CLIP-Score and a reduction of 2.87 in FID.
Authors: Zhaoxu Li, Wei An, Gaowei Guo, Longguang Wang, Yingqian Wang, Zaiping Lin
Abstract: Hyperspectral target detection (HTD) aims to identify specific materials based on spectral information in hyperspectral imagery and can detect extremely small objects, some of which occupy a smaller than one-pixel area. However, existing HTD methods are developed based on per-pixel binary classification, which limits the feature representation capability for instance-level objects. In this paper, we rethink the hyperspectral target detection from the point object detection perspective, and propose the first specialized network for hyperspectral multi-class point object detection, SpecDETR. Without the visual foundation model of the current object detection framework, SpecDETR treats each pixel in input images as a token and uses a multi-layer Transformer encoder with self-excited subpixel-scale attention modules to directly extract joint spatial-spectral features from images. During feature extraction, we introduce a self-excited mechanism to enhance object features through self-excited amplification, thereby accelerating network convergence. Additionally, SpecDETR regards point object detection as a one-to-many set prediction problem, thereby achieving a concise and efficient DETR decoder that surpasses the state-of-the-art (SOTA) DETR decoder. We develop a simulated hyperSpectral Point Object Detection benchmark termed SPOD, and for the first time, evaluate and compare the performance of current object detection networks and HTD methods on hyperspectral point object detection. Extensive experiments demonstrate that our proposed SpecDETR outperforms SOTA object detection networks and HTD methods. Our code and dataset are available at https://github.com/ZhaoxuLi123/SpecDETR.
Authors: Haocheng Dai, Sarang Joshi
Abstract: Large vision-language models (VLMs), such as CLIP, have become foundational, demonstrating remarkable success across a variety of downstream tasks. Despite their advantages, these models, akin to other foundational systems, inherit biases from the disproportionate distribution of real-world data, leading to misconceptions about the actual environment. Prevalent datasets like ImageNet are often riddled with non-causal, spurious correlations that can diminish VLM performance in scenarios where these contextual elements are absent. This study presents an investigation into how a simple linear probe can effectively distill task-specific core features from CLIP's embedding for downstream applications. Our analysis reveals that the CLIP text representations are often tainted by spurious correlations, inherited in the biased pre-training dataset. Empirical evidence suggests that relying on visual representations from CLIP, as opposed to text embedding, is more practical to refine the skewed perceptions in VLMs, emphasizing the superior utility of visual representations in overcoming embedded biases. Our codes will be available here.
Authors: Yifan Zhang, Junhui Hou
Abstract: Contrastive image-to-LiDAR knowledge transfer, commonly used for learning 3D representations with synchronized images and point clouds, often faces a self-conflict dilemma. This issue arises as contrastive losses unintentionally dissociate features of unmatched points and pixels that share semantic labels, compromising the integrity of learned representations. To overcome this, we harness Visual Foundation Models (VFMs), which have revolutionized the acquisition of pixel-level semantics, to enhance 3D representation learning. Specifically, we utilize off-the-shelf VFMs to generate semantic labels for weakly-supervised pixel-to-point contrastive distillation. Additionally, we employ von Mises-Fisher distributions to structure the feature space, ensuring semantic embeddings within the same class remain consistent across varying inputs. Furthermore, we adapt sampling probabilities of points to address imbalances in spatial distribution and category frequency, promoting comprehensive and balanced learning. Extensive experiments demonstrate that our approach mitigates the challenges posed by traditional methods and consistently surpasses existing image-to-LiDAR contrastive distillation methods in downstream tasks. The source code is available at https://github.com/Eaphan/OLIVINE.
Authors: Yisi Luo, Xile Zhao, Kai Ye, Deyu Meng
Abstract: Recently, we have witnessed the success of total variation (TV) for many imaging applications. However, traditional TV is defined on the original pixel domain, which limits its potential. In this work, we suggest a new TV regularization defined on the neural domain. Concretely, the discrete data is implicitly and continuously represented by a deep neural network (DNN), and we use the derivatives of DNN outputs w.r.t. input coordinates to capture local correlations of data. As compared with classical TV on the original domain, the proposed TV on the neural domain (termed NeurTV) enjoys the following advantages. First, NeurTV is free of discretization error induced by the discrete difference operator. Second, NeurTV is not limited to meshgrid but is suitable for both meshgrid and non-meshgrid data. Third, NeurTV can more exactly capture local correlations across data for any direction and any order of derivatives attributed to the implicit and continuous nature of neural domain. We theoretically reinterpret NeurTV under the variational approximation framework, which allows us to build the connection between NeurTV and classical TV and inspires us to develop variants (e.g., space-variant NeurTV). Extensive numerical experiments with meshgrid data (e.g., color and hyperspectral images) and non-meshgrid data (e.g., point clouds and spatial transcriptomics) showcase the effectiveness of the proposed methods.
Authors: Shizun Wang, Xingyi Yang, Qiuhong Shen, Zhenxiang Jiang, Xinchao Wang
Abstract: Recovering 4D world from monocular video is a crucial yet challenging task. Conventional methods usually rely on the assumptions of multi-view videos, known camera parameters, or static scenes. In this paper, we relax all these constraints and tackle a highly ambitious but practical task: With only one monocular video without camera parameters, we aim to recover the dynamic 3D world alongside the camera poses. To solve this, we introduce GFlow, a new framework that utilizes only 2D priors (depth and optical flow) to lift a video to a 4D scene, as a flow of 3D Gaussians through space and time. GFlow starts by segmenting the video into still and moving parts, then alternates between optimizing camera poses and the dynamics of the 3D Gaussian points. This method ensures consistency among adjacent points and smooth transitions between frames. Since dynamic scenes always continually introduce new visual content, we present prior-driven initialization and pixel-wise densification strategy for Gaussian points to integrate new content. By combining all those techniques, GFlow transcends the boundaries of 4D recovery from causal videos; it naturally enables tracking of points and segmentation of moving objects across frames. Additionally, GFlow estimates the camera poses for each frame, enabling novel view synthesis by changing camera pose. This capability facilitates extensive scene-level or object-level editing, highlighting GFlow's versatility and effectiveness. Visit our project page at: https://littlepure2333.github.io/GFlow
Authors: Minghao Guo, Bohan Wang, Pingchuan Ma, Tianyuan Zhang, Crystal Elaine Owens, Chuang Gan, Joshua B. Tenenbaum, Kaiming He, Wojciech Matusik
Abstract: We present a computational framework that transforms single images into 3D physical objects. The visual geometry of a physical object in an image is determined by three orthogonal attributes: mechanical properties, external forces, and rest-shape geometry. Existing single-view 3D reconstruction methods often overlook this underlying composition, presuming rigidity or neglecting external forces. Consequently, the reconstructed objects fail to withstand real-world physical forces, resulting in instability or undesirable deformation -- diverging from their intended designs as depicted in the image. Our optimization framework addresses this by embedding physical compatibility into the reconstruction process. We explicitly decompose the three physical attributes and link them through static equilibrium, which serves as a hard constraint, ensuring that the optimized physical shapes exhibit desired physical behaviors. Evaluations on a dataset collected from Objaverse demonstrate that our framework consistently enhances the physical realism of 3D models over existing methods. The utility of our framework extends to practical applications in dynamic simulations and 3D printing, where adherence to physical compatibility is paramount.
Authors: Junjie Zhou, Yan Shu, Bo Zhao, Boya Wu, Zhengyang Liang, Shitao Xiao, Minghao Qin, Xi Yang, Yongping Xiong, Bo Zhang, Tiejun Huang, Zheng Liu
Abstract: The evaluation of Long Video Understanding (LVU) performance poses an important but challenging research problem. Despite previous efforts, the existing video understanding benchmarks are severely constrained by several issues, especially the insufficient lengths of videos, a lack of diversity in video types and evaluation tasks, and the inappropriateness for evaluating LVU performances. To address the above problems, we propose a new benchmark called MLVU (Multi-task Long Video Understanding Benchmark) for the comprehensive and in-depth evaluation of LVU. MLVU presents the following critical values: \textit{1)} The substantial and flexible extension of video lengths, which enables the benchmark to evaluate LVU performance across a wide range of durations. \textit{2)} The inclusion of various video genres, e.g., movies, surveillance footage, egocentric videos, cartoons, game videos, etc., which reflects the models' LVU performances in different scenarios. \textit{3)} The development of diversified evaluation tasks, which enables a comprehensive examination of MLLMs' key abilities in long-video understanding. The empirical study with 23 latest MLLMs reveals significant room for improvement in today's technique, as all existing methods struggle with most of the evaluation tasks and exhibit severe performance degradation when handling longer videos. Additionally, it suggests that factors such as context length, image-understanding ability, and the choice of LLM backbone can play critical roles in future advancements. We anticipate that MLVU will advance the research of long video understanding by providing a comprehensive and in-depth analysis of MLLMs.
Authors: Scott Geng, Cheng-Yu Hsieh, Vivek Ramanujan, Matthew Wallingford, Chun-Liang Li, Pang Wei Koh, Ranjay Krishna
Abstract: Generative text-to-image models enable us to synthesize unlimited amounts of images in a controllable manner, spurring many recent efforts to train vision models with synthetic data. However, every synthetic image ultimately originates from the upstream data used to train the generator. Does the intermediate generator provide additional information over directly training on relevant parts of the upstream data? Grounding this question in the setting of image classification, we compare finetuning on task-relevant, targeted synthetic data generated by Stable Diffusion -- a generative model trained on the LAION-2B dataset -- against finetuning on targeted real images retrieved directly from LAION-2B. We show that while synthetic data can benefit some downstream tasks, it is universally matched or outperformed by real data from the simple retrieval baseline. Our analysis suggests that this underperformance is partially due to generator artifacts and inaccurate task-relevant visual details in the synthetic images. Overall, we argue that targeted retrieval is a critical baseline to consider when training with synthetic data -- a baseline that current methods do not yet surpass. We release code, data, and models at https://github.com/scottgeng00/unmet-promise.
Authors: Jiannan Wu, Muyan Zhong, Sen Xing, Zeqiang Lai, Zhaoyang Liu, Zhe Chen, Wenhai Wang, Xizhou Zhu, Lewei Lu, Tong Lu, Ping Luo, Yu Qiao, Jifeng Dai
Abstract: We present VisionLLM v2, an end-to-end generalist multimodal large model (MLLM) that unifies visual perception, understanding, and generation within a single framework. Unlike traditional MLLMs limited to text output, VisionLLM v2 significantly broadens its application scope. It excels not only in conventional visual question answering (VQA) but also in open-ended, cross-domain vision tasks such as object localization, pose estimation, and image generation and editing. To this end, we propose a new information transmission mechanism termed "super link", as a medium to connect MLLM with task-specific decoders. It not only allows flexible transmission of task information and gradient feedback between the MLLM and multiple downstream decoders but also effectively resolves training conflicts in multi-tasking scenarios. In addition, to support the diverse range of tasks, we carefully collected and combed training data from hundreds of public vision and vision-language tasks. In this way, our model can be joint-trained end-to-end on hundreds of vision language tasks and generalize to these tasks using a set of shared parameters through different user prompts, achieving performance comparable to task-specific models. We believe VisionLLM v2 will offer a new perspective on the generalization of MLLMs.
Authors: Jianan Jiang, Hao Tang, Zhilin Jiang, Weiren Yu, Di Wu
Abstract: Fine-Grained Sketch-Based Image Retrieval (FG-SBIR) aims to minimize the distance between sketches and corresponding images in the embedding space. However, scalability is hindered by the growing complexity of solutions, mainly due to the abstract nature of fine-grained sketches. In this paper, we propose an effective approach to narrow the gap between the two domains. It mainly facilitates unified mutual information sharing both intra- and inter-samples, rather than treating them as a single feature alignment problem between modalities. Specifically, our approach includes: (i) Employing dual weight-sharing networks to optimize alignment within the sketch and image domain, which also effectively mitigates model learning saturation issues. (ii) Introducing an objective optimization function based on contrastive loss to enhance the model's ability to align features in both intra- and inter-samples. (iii) Presenting a self-supervised Multi-Scale Token Recycling (MSTR) Module featured by recycling discarded patch tokens in multi-scale features, further enhancing representation capability and retrieval performance. Our framework achieves excellent results on CNN- and ViT-based backbones. Extensive experiments demonstrate its superiority over existing methods. We also introduce Cloths-V1, the first professional fashion sketch-image dataset, utilized to validate our method and will be beneficial for other applications.
Authors: Zhizhou Zhong, Yuxi Mi, Yuge Huang, Jianqing Xu, Guodong Mu, Shouhong Ding, Jingyun Zhang, Rizen Guo, Yunsheng Wu, Shuigeng Zhou
Abstract: Contemporary face recognition systems use feature templates extracted from face images to identify persons. To enhance privacy, face template protection techniques are widely employed to conceal sensitive identity and appearance information stored in the template. This paper identifies an emerging privacy attack form utilizing diffusion models that could nullify prior protection. The attack can synthesize high-quality, identity-preserving face images from templates, revealing persons' appearance. Based on studies of the diffusion model's generative capability, this paper proposes a defense by rotating templates to a noise-like distribution. This is achieved efficiently by spherically and linearly interpolating templates on their located hypersphere. This paper further proposes to group-wisely divide and drop out templates' feature dimensions, to enhance the irreversibility of rotated templates. The proposed techniques are concretized as a novel face template protection technique, SlerpFace. Extensive experiments show that SlerpFace provides satisfactory recognition accuracy and comprehensive protection against inversion and other attack forms, superior to prior arts.
Authors: Madhava Krishna, A V Subramanyam
Abstract: SimMIM is a widely used method for pretraining vision transformers using masked image modeling. However, despite its success in fine-tuning performance, it has been shown to perform sub-optimally when used for linear probing. We propose an efficient patch-wise weighting derived from keypoint features which captures the local information and provides better context during SimMIM's reconstruction phase. Our method, KAMIM, improves the top-1 linear probing accuracy from 16.12% to 33.97%, and finetuning accuracy from 76.78% to 77.3% when tested on the ImageNet-1K dataset with a ViT-B when trained for the same number of epochs. We conduct extensive testing on different datasets, keypoint extractors, and model architectures and observe that patch-wise weighting augments linear probing performance for larger pretraining datasets. We also analyze the learned representations of a ViT-B trained using KAMIM and observe that they behave similar to contrastive learning with regard to its behavior, with longer attention distances and homogenous self-attention across layers. Our code is publicly available at https://github.com/madhava20217/KAMIM.
Authors: Ezequiel Perez-Zarate, Oscar Ramos-Soto, Chunxiao Liu, Diego Oliva, Marco Perez-Cisneros
Abstract: Low-light image enhancement is an important task in computer vision, essential for improving the visibility and quality of images captured in non-optimal lighting conditions. Inadequate illumination can lead to significant information loss and poor image quality, impacting various applications such as surveillance. photography, or even autonomous driving. In this regard, automated methods have been developed to automatically adjust illumination in the image for a better visual perception. Current enhancement techniques often use specific datasets to enhance low-light images, but still present challenges when adapting to diverse real-world conditions, where illumination degradation may be localized to specific regions. To address this challenge, the Adaptive Light Enhancement Network (ALEN) is introduced, whose main approach is the use of a classification mechanism to determine whether local or global illumination enhancement is required. Subsequently, estimator networks adjust illumination based on this classification and simultaneously enhance color fidelity. ALEN integrates the Light Classification Network (LCNet) for illuminance categorization, complemented by the Single-Channel Network (SCNet), and Multi-Channel Network (MCNet) for precise estimation of illumination and color, respectively. Extensive experiments on publicly available datasets for low-light conditions were carried out to underscore ALEN's robust generalization capabilities, demonstrating superior performance in both quantitative metrics and qualitative assessments when compared to recent state-of-the-art methods. The ALEN not only enhances image quality in terms of visual perception but also represents an advancement in high-level vision tasks, such as semantic segmentation, as presented in this work. The code of this method is available at https://github.com/xingyumex/ALEN
Authors: Zhehan Zhang, Meihua Qian, Li Luo, Qianyi Gao, Xianyong Wang, Ripon Saha, Xinxin Song
Abstract: Assessing artistic creativity has long challenged researchers, with traditional methods proving time-consuming. Recent studies have applied machine learning to evaluate creativity in drawings, but not paintings. Our research addresses this gap by developing a CNN model to automatically assess the creativity of human paintings. Using a dataset of six hundred paintings by professionals and children, our model achieved 90% accuracy and faster evaluation times than human raters. This approach demonstrates the potential of machine learning in advancing artistic creativity assessment, offering a more efficient alternative to traditional methods.
Authors: Yijin Huang, Pujin Cheng, Roger Tam, Xiaoying Tang
Abstract: The success of large-scale pre-trained models has established fine-tuning as a standard method for achieving significant improvements in downstream tasks. However, fine-tuning the entire parameter set of a pre-trained model is costly. Parameter-efficient transfer learning (PETL) has recently emerged as a cost-effective alternative for adapting pre-trained models to downstream tasks. Despite its advantages, the increasing model size and input resolution present challenges for PETL, as the training memory consumption is not reduced as effectively as the parameter usage. In this paper, we introduce Fine-grained Prompt Tuning plus (FPT+), a PETL method designed for high-resolution medical image classification, which significantly reduces the training memory consumption compared to other PETL methods. FPT+ performs transfer learning by training a lightweight side network and accessing pre-trained knowledge from a large pre-trained model (LPM) through fine-grained prompts and fusion modules. Specifically, we freeze the LPM of interest and construct a learnable lightweight side network. The frozen LPM processes high-resolution images to extract fine-grained features, while the side network employs corresponding down-sampled low-resolution images to minimize the memory usage. To enable the side network to leverage pre-trained knowledge, we propose fine-grained prompts and fusion modules, which collaborate to summarize information through the LPM's intermediate activations. We evaluate FPT+ on eight medical image datasets of varying sizes, modalities, and complexities. Experimental results demonstrate that FPT+ outperforms other PETL methods, using only 1.03% of the learnable parameters and 3.18% of the memory required for fine-tuning an entire ViT-B model. Our code is available https://github.com/YijinHuang/FPT.
Authors: Bi'an Du, Lingbei Meng, Wei Hu
Abstract: Sparse-view 3D reconstruction is a major challenge in computer vision, aiming to create complete three-dimensional models from limited viewing angles. Key obstacles include: 1) a small number of input images with inconsistent information; 2) dependence on input image quality; and 3) large model parameter sizes. To tackle these issues, we propose a self-augmented two-stage Gaussian splatting framework enhanced with structural masks for sparse-view 3D reconstruction. Initially, our method generates a basic 3D Gaussian representation from sparse inputs and renders multi-view images. We then fine-tune a pre-trained 2D diffusion model to enhance these images, using them as augmented data to further optimize the 3D Gaussians. Additionally, a structural masking strategy during training enhances the model's robustness to sparse inputs and noise. Experiments on benchmarks like MipNeRF360, OmniObject3D, and OpenIllumination demonstrate that our approach achieves state-of-the-art performance in perceptual quality and multi-view consistency with sparse inputs.
Authors: Osher Rafaeli, Tal Svoray, Roni Blushtein-Livnon, Ariel Nahlieli
Abstract: This paper provides insights on the effectiveness of the zero shot, prompt-based Segment Anything Model (SAM) and its updated versions, SAM 2 and SAM 2.1, along with the non-promptable conventional neural network (CNN), for segmenting solar panels in RGB aerial imagery. The study evaluates these models across diverse lighting conditions, spatial resolutions, and prompt strategies. SAM 2 showed slight improvements over SAM, while SAM 2.1 demonstrated notable improvements, particularly in sub-optimal lighting and low resolution conditions. SAM models, when prompted by user-defined boxes, outperformed CNN in all scenarios; in particular, user-box prompts were found crucial for achieving reasonable performance in low resolution data. Additionally, under high resolution, YOLOv9 automatic prompting outperformed user-points prompting by providing reliable prompts to SAM. Under low resolution, SAM 2.1 prompted by user points showed similar performance to SAM 2.1 prompted by YOLOv9, highlighting its zero shot improvements with a single click. In high resolution with optimal lighting imagery, Eff-UNet outperformed SAMs prompted by YOLOv9, while under sub-optimal lighting conditions, Eff-UNet, and SAM 2.1 prompted by YOLOv9, had similar performance. However, SAM is more resource-intensive, and despite improved inference time of SAM 2.1, Eff-UNet is more suitable for automatic segmentation in high resolution data. This research details strengths and limitations of each model and outlines the robustness of user-prompted image segmentation models.
Authors: Zhendong Mao, Mengqi Huang, Fei Ding, Mingcong Liu, Qian He, Yongdong Zhang
Abstract: Text-to-image customization, which takes given texts and images depicting given subjects as inputs, aims to synthesize new images that align with both text semantics and subject appearance. This task provides precise control over details that text alone cannot capture and is fundamental for various real-world applications, garnering significant interest from academia and industry. Existing works follow the pseudo-word paradigm, which involves representing given subjects as pseudo-words and combining them with given texts to collectively guide the generation. However, the inherent conflict and entanglement between the pseudo-words and texts result in a dual-optimum paradox, where subject similarity and text controllability cannot be optimal simultaneously. We propose a novel real-words paradigm termed RealCustom++ that instead represents subjects as non-conflict real words, thereby disentangling subject similarity from text controllability and allowing both to be optimized simultaneously. Specifically, RealCustom++ introduces a novel "train-inference" decoupled framework: (1) During training, RealCustom++ learns the alignment between vision conditions and all real words in the text, ensuring high subject-similarity generation in open domains. This is achieved by the cross-layer cross-scale projector to robustly and finely extract subject features, and a curriculum training recipe that adapts the generated subject to diverse poses and sizes. (2) During inference, leveraging the learned general alignment, an adaptive mask guidance is proposed to only customize the generation of the specific target real word, keeping other subject-irrelevant regions uncontaminated to ensure high text-controllability in real-time.
Authors: Yihao Huang, Le Liang, Tianlin Li, Xiaojun Jia, Run Wang, Weikai Miao, Geguang Pu, Yang Liu
Abstract: In recent years, Text-to-Image (T2I) models have garnered significant attention due to their remarkable advancements. However, security concerns have emerged due to their potential to generate inappropriate or Not-Safe-For-Work (NSFW) images. In this paper, inspired by the observation that texts with different semantics can lead to similar human perceptions, we propose an LLM-driven perception-guided jailbreak method, termed PGJ. It is a black-box jailbreak method that requires no specific T2I model (model-free) and generates highly natural attack prompts. Specifically, we propose identifying a safe phrase that is similar in human perception yet inconsistent in text semantics with the target unsafe word and using it as a substitution. The experiments conducted on six open-source models and commercial online services with thousands of prompts have verified the effectiveness of PGJ.
Authors: Manuel Weber, Carly Beneke, Clyde Wheeler
Abstract: Regular measurement of carbon stock in the world's forests is critical for carbon accounting and reporting under national and international climate initiatives, and for scientific research, but has been largely limited in scalability and temporal resolution due to a lack of ground based assessments. Increasing efforts have been made to address these challenges by incorporating remotely sensed data. We present a new methodology which uses multi-sensor, multi-spectral imagery at a resolution of 10 meters and a deep learning based model which unifies the prediction of above ground biomass density (AGBD), canopy height (CH), canopy cover (CC) as well as uncertainty estimations for all three quantities. The model is trained on millions of globally sampled GEDI-L2/L4 measurements. We validate the capability of our model by deploying it over the entire globe for the year 2023 as well as annually from 2016 to 2023 over selected areas. The model achieves a mean absolute error for AGBD (CH, CC) of 26.1 Mg/ha (3.7 m, 9.9 %) and a root mean squared error of 50.6 Mg/ha (5.4 m, 15.8 %) on a globally sampled test dataset, demonstrating a significant improvement over previously published results. We also report the model performance against independently collected ground measurements published in the literature, which show a high degree of correlation across varying conditions. We further show that our pre-trained model facilitates seamless transferability to other GEDI variables due to its multi-head architecture.
Authors: Jeong Hun Yeo, Chae Won Kim, Hyunjun Kim, Hyeongseop Rha, Seunghee Han, Wen-Huang Cheng, Yong Man Ro
Abstract: Lip reading aims to predict spoken language by analyzing lip movements. Despite advancements in lip reading technologies, performance degrades when models are applied to unseen speakers due to their sensitivity to variations in visual information such as lip appearances. To address this challenge, speaker adaptive lip reading technologies have advanced by focusing on effectively adapting a lip reading model to target speakers in the visual modality. However, the effectiveness of adapting language information, such as vocabulary choice, of the target speaker has not been explored in previous works. Additionally, existing datasets for speaker adaptation have limited vocabulary sizes and pose variations, which restrict the validation of previous speaker-adaptive methods in real-world scenarios. To address these issues, we propose a novel speaker-adaptive lip reading method that adapts a pre-trained model to target speakers at both vision and language levels. Specifically, we integrate prompt tuning and the LoRA approach, applying them to a pre-trained lip reading model to effectively adapt the model to target speakers. Furthermore, to validate its effectiveness in real-world scenarios, we introduce a new dataset, VoxLRS-SA, derived from VoxCeleb2 and LRS3. It contains a vocabulary of approximately 100K words, offers diverse pose variations, and enables the validation of adaptation methods in the wild, sentence-level lip reading for the first time in English. Through various experiments, we demonstrate that the existing speaker-adaptive method also improves performance in the wild at the sentence level. Moreover, we show that the proposed method achieves larger improvements compared to the previous works.
Authors: Xianmin Chen, Peiliang Huang, Xiaoxu Feng, Dingwen Zhang, Longfei Han, Junwei Han
Abstract: Low-light image enhancement, particularly in cross-domain tasks such as mapping from the raw domain to the sRGB domain, remains a significant challenge. Many deep learning-based methods have been developed to address this issue and have shown promising results in recent years. However, single-stage methods, which attempt to unify the complex mapping across both domains, leading to limited denoising performance. In contrast, two-stage approaches typically decompose a raw image with color filter arrays (CFA) into a four-channel RGGB format before feeding it into a neural network. However, this strategy overlooks the critical role of demosaicing within the Image Signal Processing (ISP) pipeline, leading to color distortions under varying lighting conditions, especially in low-light scenarios. To address these issues, we design a novel Mamba scanning mechanism, called RAWMamba, to effectively handle raw images with different CFAs. Furthermore, we present a Retinex Decomposition Module (RDM) grounded in Retinex prior, which decouples illumination from reflectance to facilitate more effective denoising and automatic non-linear exposure correction. By bridging demosaicing and denoising, better raw image enhancement is achieved. Experimental evaluations conducted on public datasets SID and MCR demonstrate that our proposed RAWMamba achieves state-of-the-art performance on cross-domain mapping.
Authors: Ting Liu, Zunnan Xu, Yue Hu, Liangtao Shi, Zhiqiang Wang, Quanjun Yin
Abstract: Referring Expression Comprehension (REC), which aims to ground a local visual region via natural language, is a task that heavily relies on multimodal alignment. Most existing methods utilize powerful pre-trained models to transfer visual/linguistic knowledge by full fine-tuning. However, full fine-tuning the entire backbone not only breaks the rich prior knowledge embedded in the pre-training, but also incurs significant computational costs. Motivated by the recent emergence of Parameter-Efficient Transfer Learning (PETL) methods, we aim to solve the REC task in an effective and efficient manner. Directly applying these PETL methods to the REC task is inappropriate, as they lack the specific-domain abilities for precise local visual perception and visual-language alignment. Therefore, we propose a novel framework of Multimodal Prior-guided Parameter Efficient Tuning, namely MaPPER. Specifically, MaPPER comprises Dynamic Prior Adapters guided by an aligned prior, and Local Convolution Adapters to extract precise local semantics for better visual perception. Moreover, the Prior-Guided Text module is proposed to further utilize the prior for facilitating the cross-modal alignment. Experimental results on three widely-used benchmarks demonstrate that MaPPER achieves the best accuracy compared to the full fine-tuning and other PETL methods with only 1.41% tunable backbone parameters. Our code is available at https://github.com/liuting20/MaPPER.
Authors: Lukas Klein, Carsten T. L\"uth, Udo Schlegel, Till J. Bungert, Mennatallah El-Assady, Paul F. J\"ager
Abstract: Explainable AI (XAI) is a rapidly growing domain with a myriad of proposed methods as well as metrics aiming to evaluate their efficacy. However, current studies are often of limited scope, examining only a handful of XAI methods and ignoring underlying design parameters for performance, such as the model architecture or the nature of input data. Moreover, they often rely on one or a few metrics and neglect thorough validation, increasing the risk of selection bias and ignoring discrepancies among metrics. These shortcomings leave practitioners confused about which method to choose for their problem. In response, we introduce LATEC, a large-scale benchmark that critically evaluates 17 prominent XAI methods using 20 distinct metrics. We systematically incorporate vital design parameters like varied architectures and diverse input modalities, resulting in 7,560 examined combinations. Through LATEC, we showcase the high risk of conflicting metrics leading to unreliable rankings and consequently propose a more robust evaluation scheme. Further, we comprehensively evaluate various XAI methods to assist practitioners in selecting appropriate methods aligning with their needs. Curiously, the emerging top-performing method, Expected Gradients, is not examined in any relevant related study. LATEC reinforces its role in future XAI research by publicly releasing all 326k saliency maps and 378k metric scores as a (meta-)evaluation dataset. The benchmark is hosted at: https://github.com/IML-DKFZ/latec.
Authors: Ziyan Jiang, Rui Meng, Xinyi Yang, Semih Yavuz, Yingbo Zhou, Wenhu Chen
Abstract: Embedding models have been crucial in enabling various downstream tasks such as semantic similarity, information retrieval, and clustering. Recently, there has been a surge of interest in developing universal text embedding models that can generalize across tasks (e.g., MTEB). However, progress in learning universal multimodal embedding models has been relatively slow despite its importance and practicality. In this work, we aim to explore the potential for building universal embeddings capable of handling a wide range of downstream tasks. Our contributions are twofold: (1) MMEB (Massive Multimodal Embedding Benchmark), which covers 4 meta-tasks (i.e. classification, visual question answering, multimodal retrieval, and visual grounding) and 36 datasets, including 20 training and 16 evaluation datasets covering both in-distribution and out-of-distribution tasks, and (2) VLM2Vec (Vision-Language Model -> Vector), a contrastive training framework that converts any state-of-the-art vision-language model into an embedding model via training on MMEB. Unlike previous models such as CLIP and BLIP, which encodes text or images independently without any task instruction, VLM2Vec can process any combination of images and text to generate a fixed-dimensional vector based on task instructions. We build a series of VLM2Vec models on SoTA VLMs like Phi-3.5-V, LLaVA-1.6 and evaluate them on MMEB's evaluation split. Our results show that VLM2Vec achieves an absolute average improvement of 10% to 20% over existing multimodal embedding models on both in-distribution and out-of-distribution datasets in MMEB. We show that VLMs are secretly strong embedding models.
Authors: Haochen Wang, Anlin Zheng, Yucheng Zhao, Tiancai Wang, Zheng Ge, Xiangyu Zhang, Zhaoxiang Zhang
Abstract: This paper introduces reconstructive visual instruction tuning (ROSS), a family of Large Multimodal Models (LMMs) that exploit vision-centric supervision signals. In contrast to conventional visual instruction tuning approaches that exclusively supervise text outputs, ROSS prompts LMMs to supervise visual outputs via reconstructing input images. By doing so, it capitalizes on the inherent richness and detail present within input images themselves, which are often lost in pure text supervision. However, producing meaningful feedback from natural images is challenging due to the heavy spatial redundancy of visual signals. To address this issue, ROSS employs a denoising objective to reconstruct latent representations of input images, avoiding directly regressing exact raw RGB values. This intrinsic activation design inherently encourages LMMs to maintain image detail, thereby enhancing their fine-grained comprehension capabilities and reducing hallucinations. Empirically, ROSS consistently brings significant improvements across different visual encoders and language models. In comparison with extrinsic assistance state-of-the-art alternatives that aggregate multiple visual experts, ROSS delivers competitive performance with a single SigLIP visual encoder, demonstrating the efficacy of our vision-centric supervision tailored for visual outputs.
Authors: Felix Koulischer, Johannes Deleu, Gabriel Raya, Thomas Demeester, Luca Ambrogioni
Abstract: Negative Prompting (NP) is widely utilized in diffusion models, particularly in text-to-image applications, to prevent the generation of undesired features. In this paper, we show that conventional NP is limited by the assumption of a constant guidance scale, which may lead to highly suboptimal results, or even complete failure, due to the non-stationarity and state-dependence of the reverse process. Based on this analysis, we derive a principled technique called Dynamic Negative Guidance, which relies on a near-optimal time and state dependent modulation of the guidance without requiring additional training. Unlike NP, negative guidance requires estimating the posterior class probability during the denoising process, which is achieved with limited additional computational overhead by tracking the discrete Markov Chain during the generative process. We evaluate the performance of DNG class-removal on MNIST and CIFAR10, where we show that DNG leads to higher safety, preservation of class balance and image quality when compared with baseline methods. Furthermore, we show that it is possible to use DNG with Stable Diffusion to obtain more accurate and less invasive guidance than NP.
Authors: Carmel Shabalin, Israel Shenkman, Ilan Shelef, Gal Ben-Arie, Alex Geftler, Yuval Shahar
Abstract: Low back pain is the symptom that is the second most frequently reported to primary care physicians, effecting 50 to 80 percent of the population in a lifetime, resulting in multiple referrals of patients suffering from back problems, to CT and MRI scans, which are then examined by radiologists. The radiologists examining these spinal scans naturally focus on spinal pathologies and might miss other types of abnormalities, and in particular, abdominal ones, such as malignancies. Nevertheless, the patients whose spine was scanned might as well have malignant and other abdominal pathologies. Thus, clinicians have suggested the need for computerized assistance and decision support in screening spinal scans for additional abnormalities. In the current study, We have addressed the important case of detecting suspicious lesions in the adrenal glands as an example for the overall methodology we have developed. A patient CT scan is integrated from multiple slices with an axial orientation. Our method determines whether a patient has an abnormal adrenal gland, and localises the abnormality if it exists. Our method is composed of three deep learning models; each model has a different task for achieving the final goal. We call our compound method the Multi Model Graph Aggregation MMGA method. The novelty in this study is twofold. First, the use, for an important screening task, of CT scans that are originally focused and tuned for imaging the spine, which were acquired from patients with potential spinal disorders, for detection of a totally different set of abnormalities such as abdominal Adrenal glands pathologies. Second, we have built a complex pipeline architecture composed from three deep learning models that can be utilized for other organs (such as the pancreas or the kidney), or for similar applications, but using other types of imaging, such as MRI.
Authors: Xiaoliang Liu, Furao Shen, Jian Zhao
Abstract: The Segment Anything Model (SAM) is a cornerstone of image segmentation, demonstrating exceptional performance across various applications, particularly in autonomous driving and medical imaging, where precise segmentation is crucial. However, SAM is vulnerable to adversarial attacks that can significantly impair its functionality through minor input perturbations. Traditional techniques, such as FGSM and PGD, are often ineffective in segmentation tasks due to their reliance on global perturbations that overlook spatial nuances. Recent methods like Attack-SAM-K and UAD have begun to address these challenges, but they frequently depend on external cues and do not fully leverage the structural interdependencies within segmentation processes. This limitation underscores the need for a novel adversarial strategy that exploits the unique characteristics of segmentation tasks. In response, we introduce the Region-Guided Attack (RGA), designed specifically for SAM. RGA utilizes a Region-Guided Map (RGM) to manipulate segmented regions, enabling targeted perturbations that fragment large segments and expand smaller ones, resulting in erroneous outputs from SAM. Our experiments demonstrate that RGA achieves high success rates in both white-box and black-box scenarios, emphasizing the need for robust defenses against such sophisticated attacks. RGA not only reveals SAM's vulnerabilities but also lays the groundwork for developing more resilient defenses against adversarial threats in image segmentation.
Authors: Sudarshan Regmi
Abstract: Out-of-distribution (OOD) detection is critical to ensure the safe deployment of deep learning models in critical applications. Deep learning models can often misidentify OOD samples as in-distribution (ID) samples. This vulnerability worsens in the presence of spurious correlation in the training set. Likewise, in fine-grained classification settings, detection of fine-grained OOD samples becomes inherently challenging due to their high similarity to ID samples. However, current research on OOD detection has largely ignored these challenging scenarios, focusing instead on relatively easier (conventional) cases. In this work, we present a unified Approach to Spurious, fine-grained, and Conventional OOD Detection (ASCOOD). First, we propose synthesizing virtual outliers from ID data by approximating the destruction of invariant features. We identify invariant features with the pixel attribution method using the model being learned. This approach eliminates the burden of curating external OOD datasets. Then, we simultaneously incentivize ID classification and predictive uncertainty towards the virtual outliers leveraging standardized feature representation. Our approach effectively mitigates the impact of spurious correlations and encourages capturing fine-grained attributes. Extensive experiments across six datasets demonstrate the merit of ASCOOD in spurious, fine-grained, and conventional settings. The code is available at: https://github.com/sudarshanregmi/ASCOOD/
Authors: Filza Akhlaq, Alina Arshad, Muhammad Yehya Hayati, Jawwad A. Shamsi, Muhammad Burhan Khan
Abstract: Detecting mixed-critical events through computer vision is challenging due to the need for contextual understanding to assess event criticality accurately. Mixed critical events, such as fires of varying severity or traffic incidents, demand adaptable systems that can interpret context to trigger appropriate responses. This paper addresses these challenges by proposing a versatile detection system for smart city applications, offering a solution tested across traffic and fire detection scenarios. Our contributions include an analysis of detection requirements and the development of a system adaptable to diverse applications, advancing automated surveillance for smart cities.
Authors: Yuanyang Yin, Yaqi Zhao, Mingwu Zheng, Ke Lin, Jiarong Ou, Rui Chen, Victor Shea-Jay Huang, Jiahao Wang, Xin Tao, Pengfei Wan, Di Zhang, Baoqun Yin, Wentao Zhang, Kun Gai
Abstract: Achieving optimal performance of video diffusion transformers within given data and compute budget is crucial due to their high training costs. This necessitates precisely determining the optimal model size and training hyperparameters before large-scale training. While scaling laws are employed in language models to predict performance, their existence and accurate derivation in visual generation models remain underexplored. In this paper, we systematically analyze scaling laws for video diffusion transformers and confirm their presence. Moreover, we discover that, unlike language models, video diffusion models are more sensitive to learning rate and batch size, two hyperparameters often not precisely modeled. To address this, we propose a new scaling law that predicts optimal hyperparameters for any model size and compute budget. Under these optimal settings, we achieve comparable performance and reduce inference costs by 40.1% compared to conventional scaling methods, within a compute budget of 1e10 TFlops. Furthermore, we establish a more generalized and precise relationship among validation loss, any model size, and compute budget. This enables performance prediction for non-optimal model sizes, which may also be appealed under practical inference cost constraints, achieving a better trade-off.
Authors: Stuti Pathak, Prashant Kumar, Nicholus Mboga, Gunther Steenackers, Rudi Penne
Abstract: Point clouds acquired in constrained and challenging real-world settings are incomplete, non-uniformly sparse, or both. These obstacles present acute challenges for a vital task - point cloud completion. Using tools from Algebraic Topology and Persistent Homology ($\mathcal{PH}$), we demonstrate that current benchmark synthetic point clouds lack rich topological features that are important constituents of point clouds captured in realistic settings. To facilitate research in this direction, we contribute the first real-world industrial point cloud dataset for point cloud completion, RealPC - a diverse set of rich and varied point clouds, consisting of $\sim$ 40,000 pairs across 21 categories of industrial structures in railway establishments. Our benchmark results on several strong baselines reveal a striking observation - the existing methods are tailored for synthetic datasets and fail miserably in real-world settings. Building on our observation that RealPC consists of several 0 and 1-dimensional $\mathcal{PH}$-based topological features, we demonstrate the potential of integrating Homology-based topological priors into existing works. More specifically, we present how 0-dimensional $\mathcal{PH}$ priors, which extract the global topology of a complete shape in the form of a 3-D skeleton, can assist a model in generating topologically-consistent complete shapes.
Authors: Weijie Kong (refer to the report for detailed contributions), Qi Tian (refer to the report for detailed contributions), Zijian Zhang (refer to the report for detailed contributions), Rox Min (refer to the report for detailed contributions), Zuozhuo Dai (refer to the report for detailed contributions), Jin Zhou (refer to the report for detailed contributions), Jiangfeng Xiong (refer to the report for detailed contributions), Xin Li (refer to the report for detailed contributions), Bo Wu (refer to the report for detailed contributions), Jianwei Zhang (refer to the report for detailed contributions), Kathrina Wu (refer to the report for detailed contributions), Qin Lin (refer to the report for detailed contributions), Junkun Yuan (refer to the report for detailed contributions), Yanxin Long (refer to the report for detailed contributions), Aladdin Wang (refer to the report for detailed contributions), Andong Wang (refer to the report for detailed contributions), Changlin Li (refer to the report for detailed contributions), Duojun Huang (refer to the report for detailed contributions), Fang Yang (refer to the report for detailed contributions), Hao Tan (refer to the report for detailed contributions), Hongmei Wang (refer to the report for detailed contributions), Jacob Song (refer to the report for detailed contributions), Jiawang Bai (refer to the report for detailed contributions), Jianbing Wu (refer to the report for detailed contributions), Jinbao Xue (refer to the report for detailed contributions), Joey Wang (refer to the report for detailed contributions), Kai Wang (refer to the report for detailed contributions), Mengyang Liu (refer to the report for detailed contributions), Pengyu Li (refer to the report for detailed contributions), Shuai Li (refer to the report for detailed contributions), Weiyan Wang (refer to the report for detailed contributions), Wenqing Yu (refer to the report for detailed contributions), Xinchi Deng (refer to the report for detailed contributions), Yang Li (refer to the report for detailed contributions), Yi Chen (refer to the report for detailed contributions), Yutao Cui (refer to the report for detailed contributions), Yuanbo Peng (refer to the report for detailed contributions), Zhentao Yu (refer to the report for detailed contributions), Zhiyu He (refer to the report for detailed contributions), Zhiyong Xu (refer to the report for detailed contributions), Zixiang Zhou (refer to the report for detailed contributions), Zunnan Xu (refer to the report for detailed contributions), Yangyu Tao (refer to the report for detailed contributions), Qinglin Lu (refer to the report for detailed contributions), Songtao Liu (refer to the report for detailed contributions), Daquan Zhou (refer to the report for detailed contributions), Hongfa Wang (refer to the report for detailed contributions), Yong Yang (refer to the report for detailed contributions), Di Wang (refer to the report for detailed contributions), Yuhong Liu (refer to the report for detailed contributions), Jie Jiang (refer to the report for detailed contributions), Caesar Zhong (refer to the report for detailed contributions)
Abstract: Recent advancements in video generation have significantly impacted daily life for both individuals and industries. However, the leading video generation models remain closed-source, resulting in a notable performance gap between industry capabilities and those available to the public. In this report, we introduce HunyuanVideo, an innovative open-source video foundation model that demonstrates performance in video generation comparable to, or even surpassing, that of leading closed-source models. HunyuanVideo encompasses a comprehensive framework that integrates several key elements, including data curation, advanced architectural design, progressive model scaling and training, and an efficient infrastructure tailored for large-scale model training and inference. As a result, we successfully trained a video generative model with over 13 billion parameters, making it the largest among all open-source models. We conducted extensive experiments and implemented a series of targeted designs to ensure high visual quality, motion dynamics, text-video alignment, and advanced filming techniques. According to evaluations by professionals, HunyuanVideo outperforms previous state-of-the-art models, including Runway Gen-3, Luma 1.6, and three top-performing Chinese video generative models. By releasing the code for the foundation model and its applications, we aim to bridge the gap between closed-source and open-source communities. This initiative will empower individuals within the community to experiment with their ideas, fostering a more dynamic and vibrant video generation ecosystem. The code is publicly available at https://github.com/Tencent/HunyuanVideo.
Authors: Michael Schwimmbeck, Serouj Khajarian, Konstantin Holzapfel, Johannes Schmidt, Stefanie Remmele
Abstract: In the context of medical Augmented Reality (AR) applications, object tracking is a key challenge and requires a significant amount of annotation masks. As segmentation foundation models like the Segment Anything Model (SAM) begin to emerge, zero-shot segmentation requires only minimal human participation obtaining high-quality object masks. We introduce a HoloLens-Object-Labeling (HOLa) Unity and Python application based on the SAM-Track algorithm that offers fully automatic single object annotation for HoloLens 2 while requiring minimal human participation. HOLa does not have to be adjusted to a specific image appearance and could thus alleviate AR research in any application field. We evaluate HOLa for different degrees of image complexity in open liver surgery and in medical phantom experiments. Using HOLa for image annotation can increase the labeling speed by more than 500 times while providing Dice scores between 0.875 and 0.982, which are comparable to human annotators. Our code is publicly available at: https://github.com/mschwimmbeck/HOLa
Authors: Mingliang Zhai, Cheng Li, Zengyuan Guo, Ningrui Yang, Xiameng Qin, Sanyuan Zhao, Junyu Han, Ji Tao, Yuwei Wu, Yunde Jia
Abstract: The Multi-modal Large Language Models (MLLMs) with extensive world knowledge have revitalized autonomous driving, particularly in reasoning tasks within perceivable regions. However, when faced with perception-limited areas (dynamic or static occlusion regions), MLLMs struggle to effectively integrate perception ability with world knowledge for reasoning. These perception-limited regions can conceal crucial safety information, especially for vulnerable road users. In this paper, we propose a framework, which aims to improve autonomous driving performance under perceptionlimited conditions by enhancing the integration of perception capabilities and world knowledge. Specifically, we propose a plug-and-play instruction-guided interaction module that bridges modality gaps and significantly reduces the input sequence length, allowing it to adapt effectively to multi-view video inputs. Furthermore, to better integrate world knowledge with driving-related tasks, we have collected and refined a large-scale multi-modal dataset that includes 2 million natural language QA pairs, 1.7 million grounding task data. To evaluate the model's utilization of world knowledge, we introduce an object-level risk assessment dataset comprising 200K QA pairs, where the questions necessitate multi-step reasoning leveraging world knowledge for resolution. Extensive experiments validate the effectiveness of our proposed method.
Authors: Kart-Leong Lim
Abstract: Deep clustering is a recent deep learning technique which combines deep learning with traditional unsupervised clustering. At the heart of deep clustering is a loss function which penalizes samples for being an outlier from their ground truth cluster centers in the latent space. The probabilistic variant of deep clustering reformulates the loss using KL divergence. Often, the main constraint of deep clustering is the necessity of a closed form loss function to make backpropagation tractable. Inspired by deep clustering and adversarial net, we reformulate deep clustering as an adversarial net over traditional closed form KL divergence. Training deep clustering becomes a task of minimizing the encoder and maximizing the discriminator. At optimality, this method theoretically approaches the JS divergence between the distribution assumption of the encoder and the discriminator. We demonstrated the performance of our proposed method on several well cited datasets such as SVHN, USPS, MNIST and CIFAR10, achieving on-par or better performance with some of the state-of-the-art deep clustering methods.
Authors: Oishi Deb, KR Prajwal, Andrew Zisserman
Abstract: In this paper, we present a novel keypoint-based classification model designed to recognise British Sign Language (BSL) words within continuous signing sequences. Our model's performance is assessed using the BOBSL dataset, revealing that the keypoint-based approach surpasses its RGB-based counterpart in computational efficiency and memory usage. Furthermore, it offers expedited training times and demands fewer computational resources. To the best of our knowledge, this is the inaugural application of a keypoint-based model for BSL word classification, rendering direct comparisons with existing works unavailable.
Authors: Wenxuan Liu, Monde Duinkharjav, Qi Sun, Sai Qian Zhang
Abstract: Leveraging real-time eye-tracking, foveated rendering optimizes hardware efficiency and enhances visual quality virtual reality (VR). This approach leverages eye-tracking techniques to determine where the user is looking, allowing the system to render high-resolution graphics only in the foveal region-the small area of the retina where visual acuity is highest, while the peripheral view is rendered at lower resolution. However, modern deep learning-based gaze-tracking solutions often exhibit a long-tail distribution of tracking errors, which can degrade user experience and reduce the benefits of foveated rendering by causing misalignment and decreased visual quality. This paper introduces \textit{FovealNet}, an advanced AI-driven gaze tracking framework designed to optimize system performance by strategically enhancing gaze tracking accuracy. To further reduce the implementation cost of the gaze tracking algorithm, FovealNet employs an event-based cropping method that eliminates over $64.8\%$ of irrelevant pixels from the input image. Additionally, it incorporates a simple yet effective token-pruning strategy that dynamically removes tokens on the fly without compromising tracking accuracy. Finally, to support different runtime rendering configurations, we propose a system performance-aware multi-resolution training strategy, allowing the gaze tracking DNN to adapt and optimize overall system performance more effectively. Evaluation results demonstrate that FovealNet achieves at least $1.42\times$ speed up compared to previous methods and 13\% increase in perceptual quality for foveated output.
Authors: Hanfang Liang, Yizhuo Yang, Jinming Hu, Jianfei Yang, Fen Liu, Shenghai Yuan
Abstract: Compact UAV systems, while advancing delivery and surveillance, pose significant security challenges due to their small size, which hinders detection by traditional methods. This paper presents a cost-effective, unsupervised UAV detection method using spatial-temporal sequence processing to fuse multiple LiDAR scans for accurate UAV tracking in real-world scenarios. Our approach segments point clouds into foreground and background, analyzes spatial-temporal data, and employs a scoring mechanism to enhance detection accuracy. Tested on a public dataset, our solution placed 4th in the CVPR 2024 UG2+ Challenge, demonstrating its practical effectiveness. We plan to open-source all designs, code, and sample data for the research community github.com/lianghanfang/UnLiDAR-UAV-Est.
Authors: Jihao Gu, Yingyao Wang, Meng Cao, Pi Bu, Jun Song, Yancheng He, Shilong Li, Bo Zheng
Abstract: Direct Preference Optimization (DPO) has been demonstrated to be highly effective in mitigating hallucinations in Large Vision Language Models (LVLMs) by aligning their outputs more closely with human preferences. Despite the recent progress, existing methods suffer from two drawbacks: 1) Lack of scalable token-level rewards; and 2) Neglect of visual-anchored tokens. To this end, we propose a novel Token Preference Optimization model with self-calibrated rewards (dubbed as TPO), which adaptively attends to visual-correlated tokens without fine-grained annotations. Specifically, we introduce a token-level \emph{visual-anchored} \emph{reward} as the difference of the logistic distributions of generated tokens conditioned on the raw image and the corrupted one. In addition, to highlight the informative visual-anchored tokens, a visual-aware training objective is proposed to enhance more accurate token-level optimization. Extensive experimental results have manifested the state-of-the-art performance of the proposed TPO. For example, by building on top of LLAVA-1.5-7B, our TPO boosts the performance absolute improvement for hallucination benchmarks.
Authors: Eric Brouwer, Jan Erik van Woerden, Gertjan Burghouts, Matias Valdenegro-Toro, Marco Zullich
Abstract: Few-shot, fine-grained classification in computer vision poses significant challenges due to the need to differentiate subtle class distinctions with limited data. This paper presents a novel method that enhances the Contrastive Language-Image Pre-Training (CLIP) model through adaptive prompt tuning, guided by real-time visual inputs. Unlike existing techniques such as Context Optimization (CoOp) and Visual Prompt Tuning (VPT), which are constrained by static prompts or visual token reliance, the proposed approach leverages a cross-attention mechanism to dynamically refine text prompts for the image at hand. This enables an image-specific alignment of textual features with image patches extracted from the Vision Transformer, making the model more effective for datasets with high intra-class variance and low inter-class differences. The method is evaluated on several datasets, including CUBirds, Oxford Flowers, and FGVC Aircraft, showing significant performance gains over static prompt tuning approaches. To ensure these performance gains translate into trustworthy predictions, we integrate Monte-Carlo Dropout in our approach to improve the reliability of the model predictions and uncertainty estimates. This integration provides valuable insights into the model's predictive confidence, helping to identify when predictions can be trusted and when additional verification is necessary. This dynamic approach offers a robust solution, advancing the state-of-the-art for few-shot fine-grained classification.
Authors: Yatian Pang, Peng Jin, Shuo Yang, Bin Lin, Bin Zhu, Zhenyu Tang, Liuhan Chen, Francis E. H. Tay, Ser-Nam Lim, Harry Yang, Li Yuan
Abstract: Autoregressive models, built based on the Next Token Prediction (NTP) paradigm, show great potential in developing a unified framework that integrates both language and vision tasks. In this work, we rethink the NTP for autoregressive image generation and propose a novel Next Patch Prediction (NPP) paradigm. Our key idea is to group and aggregate image tokens into patch tokens containing high information density. With patch tokens as a shorter input sequence, the autoregressive model is trained to predict the next patch, thereby significantly reducing the computational cost. We further propose a multi-scale coarse-to-fine patch grouping strategy that exploits the natural hierarchical property of image data. Experiments on a diverse range of models (100M-1.4B parameters) demonstrate that the next patch prediction paradigm could reduce the training cost to around 0.6 times while improving image generation quality by up to 1.0 FID score on the ImageNet benchmark. We highlight that our method retains the original autoregressive model architecture without introducing additional trainable parameters or specifically designing a custom image tokenizer, thus ensuring flexibility and seamless adaptation to various autoregressive models for visual generation.
Authors: Huanjin Yao, Jiaxing Huang, Wenhao Wu, Jingyi Zhang, Yibo Wang, Shunyu Liu, Yingjie Wang, Yuxin Song, Haocheng Feng, Li Shen, Dacheng Tao
Abstract: In this work, we aim to develop an MLLM that understands and solves questions by learning to create each intermediate step of the reasoning involved till the final answer. To this end, we propose Collective Monte Carlo Tree Search (CoMCTS), a new learning-to-reason method for MLLMs, which introduces the concept of collective learning into ``tree search'' for effective and efficient reasoning-path searching and learning. The core idea of CoMCTS is to leverage collective knowledge from multiple models to collaboratively conjecture, search and identify effective reasoning paths toward correct answers via four iterative operations including Expansion, Simulation and Error Positioning, Backpropagation, and Selection. Using CoMCTS, we construct Mulberry-260k, a multimodal dataset with a tree of rich, explicit and well-defined reasoning nodes for each question. With Mulberry-260k, we perform collective SFT to train our model, Mulberry, a series of MLLMs with o1-like step-by-step Reasoning and Reflection capabilities. Extensive experiments demonstrate the superiority of our proposed methods on various benchmarks. Code will be available at https://github.com/HJYao00/Mulberry
Authors: Yiyuan Ge, Zhihao Chen, Ziyang Wang, Jiaju Kang, Mingya Zhang
Abstract: The development of deep learning has facilitated the application of person re-identification (ReID) technology in intelligent security. Visible-infrared person re-identification (VI-ReID) aims to match pedestrians across infrared and visible modality images enabling 24-hour surveillance. Current studies relying on unsupervised modality transformations as well as inefficient embedding constraints to bridge the spectral differences between infrared and visible images, however, limit their potential performance. To tackle the limitations of the above approaches, this paper introduces a simple yet effective Spectral Enhancement and Pseudo-anchor Guidance Network, named SEPG-Net. Specifically, we propose a more homogeneous spectral enhancement scheme based on frequency domain information and greyscale space, which avoids the information loss typically caused by inefficient modality transformations. Further, a Pseudo Anchor-guided Bidirectional Aggregation (PABA) loss is introduced to bridge local modality discrepancies while better preserving discriminative identity embeddings. Experimental results on two public benchmark datasets demonstrate the superior performance of SEPG-Net against other state-of-the-art methods. The code is available at https://github.com/1024AILab/ReID-SEPG.
Authors: Jing Li, Bo Wang
Abstract: Recent advancements in scene text spotting have focused on end-to-end methodologies that heavily rely on precise location annotations, which are often costly and labor-intensive to procure. In this study, we introduce an innovative approach that leverages only transcription annotations for training text spotting models, substantially reducing the dependency on elaborate annotation processes. Our methodology employs a query-based paradigm that facilitates the learning of implicit location features through the interaction between text queries and image embeddings. These features are later refined during the text recognition phase using an attention activation map. Addressing the challenges associated with training a weakly-supervised model from scratch, we implement a circular curriculum learning strategy to enhance model convergence. Additionally, we introduce a coarse-to-fine cross-attention localization mechanism for more accurate text instance localization. Notably, our framework supports audio-based annotation, which significantly diminishes annotation time and provides an inclusive alternative for individuals with disabilities. Our approach achieves competitive performance against existing benchmarks, demonstrating that high accuracy in text spotting can be attained without extensive location annotations.
Authors: Yijun Bei, Teng Ma, Dongxiang Zhang, Sai Wu, Kian-Lee Tan, Gang Chen
Abstract: Discovering co-movement patterns from urban-scale video data sources has emerged as an attractive topic. This task aims to identify groups of objects that travel together along a common route, which offers effective support for government agencies in enhancing smart city management. However, the previous work has made a strong assumption on the accuracy of recovered trajectories from videos and their co-movement pattern definition requires the group of objects to appear across consecutive cameras along the common route. In practice, this often leads to missing patterns if a vehicle is not correctly identified from a certain camera due to object occlusion or vehicle mis-matching. To address this challenge, we propose a relaxed definition of co-movement patterns from video data, which removes the consecutiveness requirement in the common route and accommodates a certain number of missing captured cameras for objects within the group. Moreover, a novel enumeration framework called MaxGrowth is developed to efficiently retrieve the relaxed patterns. Unlike previous filter-and-refine frameworks comprising both candidate enumeration and subsequent candidate verification procedures, MaxGrowth incurs no verification cost for the candidate patterns. It treats the co-movement pattern as an equivalent sequence of clusters, enumerating candidates with increasing sequence length while avoiding the generation of any false positives. Additionally, we also propose two effective pruning rules to efficiently filter the non-maximal patterns. Extensive experiments are conducted to validate the efficiency of MaxGrowth and the quality of its generated co-movement patterns. Our MaxGrowth runs up to two orders of magnitude faster than the baseline algorithm. It also demonstrates high accuracy in real video dataset when the trajectory recovery algorithm is not perfect.
Authors: Bharath Kumar Agnur
Abstract: This paper presents an advanced mapping system that combines drone imagery with machine learning and computer vision to overcome challenges in speed, accuracy, and adaptability across diverse terrains. By automating processes like feature detection, image matching, and stitching, the system produces seamless, high-resolution maps with minimal latency, offering strategic advantages in defense operations. Developed in Python, the system utilizes OpenCV for image processing, NumPy for efficient computations, and Concurrent[dot]futures for parallel execution. ORB (Oriented FAST and Rotated BRIEF) is employed for feature detection, while FLANN (Fast Library for Approximate Nearest Neighbors) ensures accurate keypoint matching. Homography transformations align overlapping images, resulting in distortion-free maps in real time. This automation eliminates manual intervention, enabling live updates essential in rapidly changing environments. Designed for versatility, the system performs reliably under various lighting conditions and rugged terrains, making it highly suitable for aerospace and defense applications. Testing has shown notable improvements in processing speed and accuracy compared to conventional methods, enhancing situational awareness and informed decision-making. This scalable solution leverages cutting-edge technologies to provide actionable, reliable data for mission-critical operations.
Authors: Daiheng Gao, Shilin Lu, Shaw Walters, Wenbo Zhou, Jiaming Chu, Jie Zhang, Bang Zhang, Mengxi Jia, Jian Zhao, Zhaoxin Fan, Weiming Zhang
Abstract: Removing unwanted concepts from large-scale text-to-image (T2I) diffusion models while maintaining their overall generative quality remains an open challenge. This difficulty is especially pronounced in emerging paradigms, such as Stable Diffusion (SD) v3 and Flux, which incorporate flow matching and transformer-based architectures. These advancements limit the transferability of existing concept-erasure techniques that were originally designed for the previous T2I paradigm (e.g., SD v1.4). In this work, we introduce EraseAnything, the first method specifically developed to address concept erasure within the latest flow-based T2I framework. We formulate concept erasure as a bi-level optimization problem, employing LoRA-based parameter tuning and an attention map regularizer to selectively suppress undesirable activations. Furthermore, we propose a self-contrastive learning strategy to ensure that removing unwanted concepts does not inadvertently harm performance on unrelated ones. Experimental results demonstrate that EraseAnything successfully fills the research gap left by earlier methods in this new T2I paradigm, achieving state-of-the-art performance across a wide range of concept erasure tasks.
Authors: Skanda Bharadwaj, Robert Collins, Yanxi Liu
Abstract: Classical approaches to Vanishing Point Detection (VPD) rely solely on the presence of explicit straight lines in images, while recent supervised deep learning approaches need labeled datasets for training. We propose an alternative unsupervised approach: Recurrence-based Vanishing Point Detection (R-VPD) that uses implicit lines discovered from recurring correspondences in addition to explicit lines. Furthermore, we contribute two Recurring-Pattern-for-Vanishing-Point (RPVP) datasets: 1) a Synthetic Image dataset with 3,200 ground truth vanishing points and camera parameters, and 2) a Real-World Image dataset with 1,400 human annotated vanishing points. We compare our method with two classical methods and two state-of-the-art deep learning-based VPD methods. We demonstrate that our unsupervised approach outperforms all the methods on the synthetic images dataset, outperforms the classical methods, and is on par with the supervised learning approaches on real-world images.
Authors: Langyu Wang, Bingke Zhu, Yingying Chen, Jinqiao Wang
Abstract: Audio-visual video parsing focuses on classifying videos through weak labels while identifying events as either visible, audible, or both, alongside their respective temporal boundaries. Many methods ignore that different modalities often lack alignment, thereby introducing extra noise during modal interaction. In this work, we introduce a Learning Interaction method for Non-aligned Knowledge (LINK), designed to equilibrate the contributions of distinct modalities by dynamically adjusting their input during event prediction. Additionally, we leverage the semantic information of pseudo-labels as a priori knowledge to mitigate noise from other modalities. Our experimental findings demonstrate that our model outperforms existing methods on the LLP dataset.
Authors: Qingyan Bai, Hao Ouyang, Yinghao Xu, Qiuyu Wang, Ceyuan Yang, Ka Leong Cheng, Yujun Shen, Qifeng Chen
Abstract: As a verified need, consistent editing across in-the-wild images remains a technical challenge arising from various unmanageable factors, like object poses, lighting conditions, and photography environments. Edicho steps in with a training-free solution based on diffusion models, featuring a fundamental design principle of using explicit image correspondence to direct editing. Specifically, the key components include an attention manipulation module and a carefully refined classifier-free guidance (CFG) denoising strategy, both of which take into account the pre-estimated correspondence. Such an inference-time algorithm enjoys a plug-and-play nature and is compatible to most diffusion-based editing methods, such as ControlNet and BrushNet. Extensive results demonstrate the efficacy of Edicho in consistent cross-image editing under diverse settings. We will release the code to facilitate future studies.
Authors: Yuanbo Yang, Jiahao Shao, Xinyang Li, Yujun Shen, Andreas Geiger, Yiyi Liao
Abstract: In this work, we introduce Prometheus, a 3D-aware latent diffusion model for text-to-3D generation at both object and scene levels in seconds. We formulate 3D scene generation as multi-view, feed-forward, pixel-aligned 3D Gaussian generation within the latent diffusion paradigm. To ensure generalizability, we build our model upon pre-trained text-to-image generation model with only minimal adjustments, and further train it using a large number of images from both single-view and multi-view datasets. Furthermore, we introduce an RGB-D latent space into 3D Gaussian generation to disentangle appearance and geometry information, enabling efficient feed-forward generation of 3D Gaussians with better fidelity and geometry. Extensive experimental results demonstrate the effectiveness of our method in both feed-forward 3D Gaussian reconstruction and text-to-3D generation. Project page: https://freemty.github.io/project-prometheus/
Authors: M. Alex O. Vasilescu
Abstract: We derive a set of causal deep neural networks whose architectures are a consequence of tensor (multilinear) factor analysis, a framework that facilitates forward and inverse causal inference. Forward causal questions are addressed with a neural architecture composed of causal capsules and a tensor transformer. Causal capsules compute a set of invariant causal factor representations, whose interactions are governed by a tensor transformation. Inverse causal questions are addressed with a neural network that implements the multilinear projection algorithm. The architecture reverses the order of the operations of a forward neural network and estimates the causes of effects. As an alternative to aggressive bottleneck dimension reduction or regularized regression that may camouflage an inherently underdetermined inverse problem, we prescribe modeling different aspects of the mechanism of data formation with piecewise tensor models whose multilinear projections produce multiple candidate solutions. Our forward and inverse questions may be addressed with shallow architectures, but for computationally scalable solutions, we derive a set of deep neural networks by taking advantage of block algebra. An interleaved kernel hierarchy results in a doubly non-linear tensor factor models. The causal neural networks that are a consequence of tensor factor analysis are data agnostic, but are illustrated with facial images. Sequential, parallel and asynchronous parallel computation strategies are described.
Authors: Haiyang Chen, Juan Gao, Zhuo Chen, Chenhao Gao, Sirui Huo, Meng Jiang, Jun Pu, Chenxi Hu
Abstract: Background: Current mainstream cardiovascular magnetic resonance-feature tracking (CMR-FT) methods, including optical flow and pairwise registration, often suffer from the drift effect caused by accumulative tracking errors. Here, we developed a CMR-FT method based on deformable groupwise registration with a locally low-rank (LLR) dissimilarity metric to improve myocardial tracking and strain estimation accuracy. Methods: The proposed method, Groupwise-LLR, performs feature tracking by iteratively updating the entire displacement field across all cardiac phases to minimize the sum of the patchwise signal ranks of the deformed movie. The method was compared with alternative CMR-FT methods including the Farneback optical flow, a sequentially pairwise registration method, and a global low rankness-based groupwise registration method via a simulated dataset (n = 20), a public cine data set (n = 100), and an in-house tagging-MRI patient dataset (n = 16). The proposed method was also compared with two general groupwise registration methods, nD+t B-Splines and pTVreg, in simulations and in vivo tracking. Results: On the simulated dataset, Groupwise-LLR achieved the lowest point tracking errors and voxelwise/global strain errors. On the public dataset, Groupwise-LLR achieved the lowest contour tracking errors, reduced the drift effect in late-diastole, and preserved similar inter-observer reproducibility as the alternative methods. On the patient dataset, Groupwise-LLR correlated better with tagging-MRI for radial strains than the other CMR-FT methods in multiple myocardial segments and levels. Conclusions: The proposed Groupwise-LLR reduces the drift effect and provides more accurate myocardial tracking and strain estimation than the alternative methods. The method may thus facilitate a more accurate estimation of myocardial strains for clinical assessments of cardiac function.
Authors: Noga Kertes, Yael Zaffrani-Reznikov, Onur Afacan, Sila Kurugol, Simon K. Warfield, Moti Freiman
Abstract: Quantitative analysis of pseudo-diffusion in diffusion-weighted magnetic resonance imaging (DWI) data shows potential for assessing fetal lung maturation and generating valuable imaging biomarkers. Yet, the clinical utility of DWI data is hindered by unavoidable fetal motion during acquisition. We present IVIM-morph, a self-supervised deep neural network model for motion-corrected quantitative analysis of DWI data using the Intra-voxel Incoherent Motion (IVIM) model. IVIM-morph combines two sub-networks, a registration sub-network, and an IVIM model fitting sub-network, enabling simultaneous estimation of IVIM model parameters and motion. To promote physically plausible image registration, we introduce a biophysically informed loss function that effectively balances registration and model-fitting quality. We validated the efficacy of IVIM-morph by establishing a correlation between the predicted IVIM model parameters of the lung and gestational age (GA) using fetal DWI data of 39 subjects. IVIM-morph exhibited a notably improved correlation with gestational age (GA) when performing in-vivo quantitative analysis of fetal lung DWI data during the canalicular phase. IVIM-morph shows potential in developing valuable biomarkers for non-invasive assessment of fetal lung maturity with DWI data. Moreover, its adaptability opens the door to potential applications in other clinical contexts where motion compensation is essential for quantitative DWI analysis. The IVIM-morph code is readily available at: https://github.com/TechnionComputationalMRILab/qDWI-Morph.
URLs: https://github.com/TechnionComputationalMRILab/qDWI-Morph.
Authors: Kai Wang, Dongwen Tang, Boya Zeng, Yida Yin, Zhaopan Xu, Yukun Zhou, Zelin Zang, Trevor Darrell, Zhuang Liu, Yang You
Abstract: Diffusion models have achieved remarkable success in image and video generation. In this work, we demonstrate that diffusion models can also \textit{generate high-performing neural network parameters}. Our approach is simple, utilizing an autoencoder and a diffusion model. The autoencoder extracts latent representations of a subset of the trained neural network parameters. Next, a diffusion model is trained to synthesize these latent representations from random noise. This model then generates new representations, which are passed through the autoencoder's decoder to produce new subsets of high-performing network parameters. Across various architectures and datasets, our approach consistently generates models with comparable or improved performance over trained networks, with minimal additional cost. Notably, we empirically find that the generated models are not memorizing the trained ones. Our results encourage more exploration into the versatile use of diffusion models. Our code is available \href{https://github.com/NUS-HPC-AI-Lab/Neural-Network-Diffusion}{here}.
URLs: https://github.com/NUS-HPC-AI-Lab/Neural-Network-Diffusion
Authors: Philipp Hess, Michael Aich, Baoxiang Pan, Niklas Boers
Abstract: Accurate and high-resolution Earth system model (ESM) simulations are essential to assess the ecological and socio-economic impacts of anthropogenic climate change, but are computationally too expensive to be run at sufficiently high spatial resolution. Recent machine learning approaches have shown promising results in downscaling ESM simulations, outperforming state-of-the-art statistical approaches. However, existing methods require computationally costly retraining for each ESM and extrapolate poorly to climates unseen during training. We address these shortcomings by learning a consistency model (CM) that efficiently and accurately downscales arbitrary ESM simulations without retraining in a zero-shot manner. Our approach yields probabilistic downscaled fields at a resolution only limited by the observational reference data. We show that the CM outperforms state-of-the-art diffusion models at a fraction of computational cost while maintaining high controllability on the downscaling task. Further, our method generalizes to climate states unseen during training without explicitly formulated physical constraints.
Authors: Ziqiang Shi, Rujie Liu
Abstract: Diffusion generative modelling (DGM) based on stochastic differential equations (SDEs) with score matching has achieved unprecedented results in data generation. In this paper, we propose a novel fast high-quality generative modelling method based on high-order Langevin dynamics (HOLD) with score matching. This motive is proved by third-order Langevin dynamics. By augmenting the previous SDEs, e.g. variance exploding or variance preserving SDEs for single-data variable processes, HOLD can simultaneously model position, velocity, and acceleration, thereby improving the quality and speed of the data generation at the same time. HOLD is composed of one Ornstein-Uhlenbeck process and two Hamiltonians, which reduce the mixing time by two orders of magnitude. Empirical experiments for unconditional image generation on the public data set CIFAR-10 and CelebA-HQ show that the effect is significant in both Frechet inception distance (FID) and negative log-likelihood, and achieves the state-of-the-art FID of 1.85 on CIFAR-10.
Authors: Yuxuan Qi, Li Lin, Jiajun Wang, Bin Zhang, Jingya Zhang
Abstract: Accurate tumor segmentation in PET/CT images is crucial for computer-aided cancer diagnosis and treatment. The primary challenge lies in effectively integrating the complementary information from PET and CT images. In clinical settings, the quality of PET and CT images often varies significantly, leading to uncertainty in the modality information extracted by networks. To address this challenge, we propose a novel Multi-modal Evidential Fusion Network (MEFN), which consists of two core stages: Cross-Modal Feature Learning (CFL) and Multi-modal Trustworthy Fusion (MTF). The CFL stage aligns features across different modalities and learns more robust feature representations, thereby alleviating the negative effects of domain gap. The MTF stage utilizes mutual attention mechanisms and an uncertainty calibrator to fuse modality features based on modality uncertainty and then fuse the segmentation results under the guidance of Dempster-Shafer Theory. Besides, a new uncertainty perceptual loss is introduced to force the model focusing on uncertain features and hence improve its ability to extract trusted modality information. Extensive comparative experiments are conducted on two publicly available PET/CT datasets to evaluate the performance of our proposed method whose results demonstrate that our MEFN significantly outperforms state-of-the-art methods with improvements of 3.10% and 3.23% in DSC scores on the AutoPET dataset and the Hecktor dataset, respectively. More importantly, our model can provide radiologists with credible uncertainty of the segmentation results for their decision in accepting or rejecting the automatic segmentation results, which is particularly important for clinical applications. Our code will be available at https://github.com/QPaws/MEFN.
Authors: Malte T\"olle, Philipp Garthe, Clemens Scherer, Jan Moritz Seliger, Andreas Leha, Nina Kr\"uger, Stefan Simm, Simon Martin, Sebastian Eble, Halvar Kelm, Moritz Bednorz, Florian Andr\'e, Peter Bannas, Gerhard Diller, Norbert Frey, Stefan Gro{\ss}, Anja Hennemuth, Lars Kaderali, Alexander Meyer, Eike Nagel, Stefan Orwat, Moritz Seiffert, Tim Friede, Tim Seidler, Sandy Engelhardt
Abstract: Federated learning is a renowned technique for utilizing decentralized data while preserving privacy. However, real-world applications often face challenges like partially labeled datasets, where only a few locations have certain expert annotations, leaving large portions of unlabeled data unused. Leveraging these could enhance transformer architectures ability in regimes with small and diversely annotated sets. We conduct the largest federated cardiac CT analysis to date (n=8,104) in a real-world setting across eight hospitals. Our two-step semi-supervised strategy distills knowledge from task-specific CNNs into a transformer. First, CNNs predict on unlabeled data per label type and then the transformer learns from these predictions with label-specific heads. This improves predictive accuracy and enables simultaneous learning of all partial labels across the federation, and outperforms UNet-based models in generalizability on downstream tasks. Code and model weights are made openly available for leveraging future cardiac CT analysis.
Authors: Shun Zou, Zhuo Zhang, Guangwei Gao
Abstract: Optical Coherence Tomography Angiography (OCTA) is a crucial imaging technique for visualizing retinal vasculature and diagnosing eye diseases such as diabetic retinopathy and glaucoma. However, precise segmentation of OCTA vasculature remains challenging due to the multi-scale vessel structures and noise from poor image quality and eye lesions. In this study, we proposed OCTAMamba, a novel U-shaped network based on the Mamba architecture, designed to segment vasculature in OCTA accurately. OCTAMamba integrates a Quad Stream Efficient Mining Embedding Module for local feature extraction, a Multi-Scale Dilated Asymmetric Convolution Module to capture multi-scale vasculature, and a Focused Feature Recalibration Module to filter noise and highlight target areas. Our method achieves efficient global modeling and local feature extraction while maintaining linear complexity, making it suitable for low-computation medical applications. Extensive experiments on the OCTA 3M, OCTA 6M, and ROSSA datasets demonstrated that OCTAMamba outperforms state-of-the-art methods, providing a new reference for efficient OCTA segmentation. Code is available at https://github.com/zs1314/OCTAMamba
Authors: Jian Wang, Razieh Faghihpirayesh, Danny Joca, Polina Golland, Ali Gholipour
Abstract: In this paper, we introduce a Universal Motion Correction (UniMo) framework, leveraging deep neural networks to tackle the challenges of motion correction across diverse imaging modalities. Our approach employs advanced neural network architectures with equivariant filters, overcoming the limitations of current models that require iterative inference or retraining for new image modalities. UniMo enables one-time training on a single modality while maintaining high stability and adaptability for inference across multiple unseen image modalities. We developed a joint learning framework that integrates multimodal knowledge from both shape and images that faithfully improve motion correction accuracy despite image appearance variations. UniMo features a geometric deformation augmenter that enhances the robustness of global motion correction by addressing any local deformations whether they are caused by object deformations or geometric distortions, and also generates augmented data to improve the training process. Our experimental results, conducted on various datasets with four different image modalities, demonstrate that UniMo surpasses existing motion correction methods in terms of accuracy. By offering a comprehensive solution to motion correction, UniMo marks a significant advancement in medical imaging, especially in challenging applications with wide ranges of motion, such as fetal imaging. The code for this work is available online, https://github.com/IntelligentImaging/UNIMO/.
Authors: Naren Sengodan
Abstract: Breast cancer diagnosis relies heavily on histopathology image classification. This study proposes a novel approach leveraging Hybrid EfficientNet models integrated with advanced attention mechanisms (CB and deformable attention) to enhance feature extraction and focus on relevant tissue regions. Evaluating on the BreakHis dataset across multiple magnification scales (40X, 100X, 200X, 400X), we achieve state-of-the-art performance with EfficientNetV2-XL and CB, reaching 98.96% accuracy and 98.31% F1-score at 400X. Integration of CLAHE preprocessing and optimized computational efficiency demonstrates suitability for real-time clinical deployment.
Authors: Yu Zhang, Shuang Li, Yibing Wang, Yu Sun, Wenyi Xiang
Abstract: Photoacoustic imaging (PAI) suffers from inherent limitations that can degrade the quality of reconstructed results, such as noise, artifacts and incomplete data acquisition caused by sparse sampling or partial array detection. In this study, we proposed a new optimization method for both two-dimensional (2D) and three-dimensional (3D) PAI reconstruction results, called the regularized iteration method with shape prior. The shape prior is a probability matrix derived from the reconstruction results of multiple sets of random partial array signals in a computational imaging system using any reconstruction algorithm, such as Delay-and-Sum (DAS) and Back-Projection (BP). In the probability matrix, high-probability locations indicate high consistency among multiple reconstruction results at those positions, suggesting a high likelihood of representing the true imaging results. In contrast, low-probability locations indicate higher randomness, leaning more towards noise or artifacts. As a shape prior, this probability matrix guides the iteration and regularization of the entire array signal reconstruction results using the original reconstruction algorithm (the same algorithm for processing random partial array signals). The method takes advantage of the property that the similarity of the object to be imitated is higher than that of noise or artifact in the results reconstructed by multiple sets of random partial array signals of the entire imaging system. The probability matrix is taken as a prerequisite for improving the original reconstruction results, and the optimizer is used to further iterate the imaging results to remove noise and artifacts and improve the imaging fidelity. Especially in the case involving sparse view which brings more artifacts, the effect is remarkable. Simulation and real experiments have both demonstrated the superiority of this method.
Authors: Riccardo Fosco Gramaccioni, Christian Marinoni, Emilian Postolache, Marco Comunit\`a, Luca Cosmo, Joshua D. Reiss, Danilo Comminiello
Abstract: Sound designers and Foley artists usually sonorize a scene, such as from a movie or video game, by manually annotating and sonorizing each action of interest in the video. In our case, the intent is to leave full creative control to sound designers with a tool that allows them to bypass the more repetitive parts of their work, thus being able to focus on the creative aspects of sound production. We achieve this presenting Stable-V2A, a two-stage model consisting of: an RMS-Mapper that estimates an envelope representative of the audio characteristics associated with the input video; and Stable-Foley, a diffusion model based on Stable Audio Open that generates audio semantically and temporally aligned with the target video. Temporal alignment is guaranteed by the use of the envelope as a ControlNet input, while semantic alignment is achieved through the use of sound representations chosen by the designer as cross-attention conditioning of the diffusion process. We train and test our model on Greatest Hits, a dataset commonly used to evaluate V2A models. In addition, to test our model on a case study of interest, we introduce Walking The Maps, a dataset of videos extracted from video games depicting animated characters walking in different locations. Samples and code available on our demo page at https://ispamm.github.io/Stable-V2A.
Authors: Uzoamaka Ezeakunne, Chrisantus Eze, Xiuwen Liu
Abstract: Despite the progress made in deepfake detection research, recent studies have shown that biases in the training data for these detectors can result in varying levels of performance across different demographic groups, such as race and gender. These disparities can lead to certain groups being unfairly targeted or excluded. Traditional methods often rely on fair loss functions to address these issues, but they under-perform when applied to unseen datasets, hence, fairness generalization remains a challenge. In this work, we propose a data-driven framework for tackling the fairness generalization problem in deepfake detection by leveraging synthetic datasets and model optimization. Our approach focuses on generating and utilizing synthetic data to enhance fairness across diverse demographic groups. By creating a diverse set of synthetic samples that represent various demographic groups, we ensure that our model is trained on a balanced and representative dataset. This approach allows us to generalize fairness more effectively across different domains. We employ a comprehensive strategy that leverages synthetic data, a loss sharpness-aware optimization pipeline, and a multi-task learning framework to create a more equitable training environment, which helps maintain fairness across both intra-dataset and cross-dataset evaluations. Extensive experiments on benchmark deepfake detection datasets demonstrate the efficacy of our approach, surpassing state-of-the-art approaches in preserving fairness during cross-dataset evaluation. Our results highlight the potential of synthetic datasets in achieving fairness generalization, providing a robust solution for the challenges faced in deepfake detection.
Authors: Shengbo Tan, Rundong Xue, Shipeng Luo, Zeyu Zhang, Xinran Wang, Lei Zhang, Daji Ergu, Zhang Yi, Yang Zhao, Ying Cai
Abstract: Hepatic vessels in computed tomography scans often suffer from image fragmentation and noise interference, making it difficult to maintain vessel integrity and posing significant challenges for vessel segmentation. To address this issue, we propose an innovative model: SegKAN. First, we improve the conventional embedding module by adopting a novel convolutional network structure for image embedding, which smooths out image noise and prevents issues such as gradient explosion in subsequent stages. Next, we transform the spatial relationships between Patch blocks into temporal relationships to solve the problem of capturing positional relationships between Patch blocks in traditional Vision Transformer models. We conducted experiments on a Hepatic vessel dataset, and compared to the existing state-of-the-art model, the Dice score improved by 1.78%. These results demonstrate that the proposed new structure effectively enhances the segmentation performance of high-resolution extended objects. Code will be available at https://github.com/goblin327/SegKAN
Authors: Supriya Manna, Niladri Sett
Abstract: Deep learning's preponderance across scientific domains has reshaped high-stakes decision-making, making it essential to follow rigorous operational frameworks that include both Right-to-Privacy (RTP) and Right-to-Explanation (RTE). This paper examines the complexities of combining these two requirements. For RTP, we focus on `Differential privacy' (DP), which is considered the current \textit{gold standard} for privacy-preserving machine learning due to its strong quantitative guarantee of privacy. For RTE, we focus on post-hoc explainers: they are the \textit{go-to} option for model auditing as they operate independently of model training. We formally investigate DP models and various commonly-used post-hoc explainers: how to evaluate these explainers subject to RTP, and analyze the intrinsic interactions between DP models and these explainers. Furthermore, our work throws light on how RTP and RTE can be effectively combined in high-stakes applications. Our study concludes by outlining an industrial software pipeline, with the example of a wildly used use-case, that respects both RTP and RTE requirements.