Multi-Agent collaboration addresses inherent limitations of individual agent systems, including limited sensing range and occlusion-induced blind spots. Despite significant progress, persistent challenges such as constrained communication bandwidth and under-explored subsequent extensions still hinder real-time deployment and further developments of collaborative autonomous driving systems. In this work, we propose ZeRCP, a unified communication-efficient framework that bridges collaborative perception with future scene prediction. Specifically, (i) we devise a plug-and-play request-free spatial filtering module (ZeroR) that eliminates the reliance on request maps while preserving inter-agent spatial complementarity modeling. This approach further reduce communication latency and bandwidth consumptions. (ii) We design a multi-scale pyramidal prediction network anchored by a novel Spatial-Temporal Deformable Attention (STDA) module, extending frame-wise detection to multi-frame predictions. This method adeptly models spatiotemporal dynamics without relying on auto-regressive recursion. We evaluate our method on a large-scale dataset in challenging semantic segmentation and scene prediction tasks. Extensive experiments demonstrate the superiority and effectiveness of ZeRCP in bandwidth-constrained collaboration scenarios and spatiotemporal prediction applications.

Object detection methods have evolved from closed-set to open-set paradigms over the years. Current open-set object detectors, however, remain constrained by their exclusive reliance on positive indicators based on given prompts like text descriptions or visual exemplars. This positive-only paradigm experiences consistent vulnerability to visually similar but semantically different distractors. We propose T-Rex-Omni, a novel framework that addresses this limitation by incorporating negative visual prompts to negate hard negative distractors. Specifically, we first introduce a unified visual prompt encoder that jointly processes positive and negative visual prompts. Next, a training-free Negating Negative Computing (NNC) module is proposed to dynamically suppress negative responses during the probability computing stage. To further boost performance through fine-tuning, our Negating Negative Hinge (NNH) loss enforces discriminative margins between positive and negative embeddings. T-Rex-Omni supports flexible deployment in both positive-only and joint positive-negative inference modes, accommodating either user-specified or automatically generated negative examples. Extensive experiments demonstrate remarkable zero-shot detection performance, significantly narrowing the performance gap between visual-prompted and text-prompted methods while showing particular strength in long-tailed scenarios (51.2 AP_r on LVIS-minival). This work establishes negative prompts as a crucial new dimension for advancing open-set visual recognition systems.

Zhen Yang, Mingyang Zhang, Feng Chen, Ganggui Ding, Liang Hou, Xin Tao, Ying-Cong Chen. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Video generation has witnessed significant advancements, yet evaluating these models remains a challenge. A comprehensive evaluation benchmark for video generation is indispensable for two reasons: 1) Existing metrics do not fully align with human perceptions; 2) An ideal evaluation system should provide insights to inform future developments of video generation. To this end, we present VBench++, a comprehensive benchmark suite that dissects "video generation quality" into specific, hierarchical, and disentangled dimensions, each with tailored prompts and evaluation methods. VBench++ has several appealing properties: 1) Comprehensive Dimensions: VBench++ comprises 16 dimensions in text-to-video generation (e.g., subject identity inconsistency, motion smoothness, temporal flickering, and spatial relationship, etc). The evaluation metrics with fine-grained levels reveal individual models' strengths and weaknesses. 2) Human Alignment: We also provide a dataset of human preference annotations to validate our benchmarks' alignment with human perception, for each evaluation dimension respectively. 3) Valuable Insights: We look into current models' ability across various evaluation dimensions, and various content types. We also investigate the gaps between video and image generation models. 4) Versatile Benchmarking: VBench++ is designed to evaluate a wide range of video generation tasks, including text-to-video and image-to-video. We introduce a high-quality Image Suite with an adaptive aspect ratio to enable fair evaluations across different image-to-video generation settings. Beyond assessing technical quality, VBench++ evaluates the trustworthiness of video generative models, providing a more holistic view of model performance. 5) Full Open-Sourcing: We fully open-source VBench++, including all prompts, the Image Suite, evaluation methods, generated videos, and human preference annotations.

We introduce Iris3D, a novel 3D content generation system that generates vivid textures and detailed 3D shapes while preserving the input information. Our system integrates a Multi-View Large Reconstruction Model (MVLRM [Li et al. 2023b ]) to generate a coarse 3D mesh and introduces a novel optimization scheme called Synchronized Diffusion Distillation (SDD) for refinement. Unlike previous refined methods based on Score Distillation Sampling (SDS), which suffer from unstable optimization and geometric over-smoothing due to ambiguities across different views and modalities, our method effectively distills consistent multi-view and multi-modal priors from 2D diffusion models in a training-free manner. This enables robust optimization of 3D representations. Additionally, because SDD is training-free, it preserves the diffusion’s prior knowledge and mitigates potential degradation. This characteristic makes it highly compatible with advanced 2D diffusion techniques like IP-Adapters and ControlNet, allowing for more controllable 3D generation with additional conditioning signals. Experiments demonstrate that our method produces high-quality 3D results with plausible textures and intricate geometric details.

Detecting and localizing objects in 3D space using multiple cameras, known as Multi-Camera 3D Object Detection (MC3D-Det), has gained prominence with the advent of bird's-eye view (BEV) approaches. However, these methods often struggle with the serious domain gaps caused by various viewpoints and environments between the training and testing domains. To address this challenge, we propose a novel framework that aligns 3D detection with 2D camera plane results by perspective rendering, thus achieving consistent and accurate results when facing serious domain shifts. Our approach consists of two main steps in both source and target domains: 1) rendering diverse view maps from BEV features by leveraging implicit foreground volumes and 2) rectifying the perspective bias of these maps. This design promotes the learning of perspective- and context-independent features, crucial for accurate object detection across varying viewpoints, camera parameters, and environmental conditions. Notably, our model-agnostic approach preserves the original network structure without incurring additional inference costs, facilitating seamless integration across various models and simplifying deployment. Worth noting is that our approach achieves satisfactory results in real data when trained only with virtual datasets, eliminating the need for real scene annotations. Experimental results on both Domain Generalization (DG) and Unsupervised Domain Adaptation (UDA) demonstrate its effectiveness.

Poster design is a critical medium for visual communication. Prior work has explored automatic poster design using deep learning techniques, but these approaches lack text accuracy, user customization, and aesthetic appeal, limiting their applicability in artistic domains such as movies and exhibitions, where both clear content delivery and visual impact are essential. To address these limitations, we present POSTA: a modular framework powered by diffusion models and multimodal large language models (MLLMs) for customized artistic poster generation. The framework consists of three modules. Background Diffusion creates a themed background based on user input. Design MLLM then generates layout and typography elements that align with and complement the background style. Finally, to enhance the poster’s aesthetic appeal, ArtText Diffusion applies additional stylization to key text elements. The final result is a visually cohesive and appealing poster, with a fully modular process that allows for complete customization. To train our models, we develop the PosterArt dataset, comprising high-quality artistic posters annotated with layout, typography, and pixel-level stylized text segmentation. Our comprehensive experimental analysis demonstrates POSTA’s exceptional controllability and design diversity, outperforming existing models in both text accuracy and aesthetic quality.

Rendering and inverse rendering are pivotal tasks in both computer vision and graphics. The rendering equation is the core of the two tasks, as an ideal conditional distribution transfer function from intrinsic properties to RGB images. Despite achieving promising results of existing rendering methods, they merely approximate the ideal estimation for a specific scene and come with a high computational cost. Additionally, the inverse conditional distribution transfer is intractable due to the inherent ambiguity. To address these challenges, we propose a data-driven method that jointly models rendering and inverse rendering as two conditional generation tasks within a single diffusion framework. Inspired by UniDiffuser, we utilize two distinct time schedules to model both tasks, and with a tailored dual streaming module, we achieve cross-conditioning of two pre-trained diffusion models. This unified approach, named Uni-Renderer, allows the two processes to facilitate each other through a cycle-consistent constrain, mitigating ambiguity by enforcing consistency between intrinsic properties and rendered images. Combined with a meticulously prepared dataset, our method effectively decomposition of intrinsic properties and demonstrating a strong capability to recognize changes during rendering.

Diffusion models have achieved great success in generating 2D images. However, the quality and generaliz-ability of 3D content generation remain limited. State- of-the-art methods often require large-scale 3D assets for training, which are challenging to collect. In this work, we introduce Kiss3DGen (Keep It Simple and Straightforward in 3D Generation), an efficient framework for generating, editing, and enhancing 3D objects by repurposing a well-trained 2D image diffusion model for 3D generation. Specifically, we fine-tune a diffusion model to generate "3D Bundle Image", a tiled representation composed of multi-view images and their corresponding normal maps. The normal maps are then used to reconstruct a 3D mesh, and the multi-view images provide texture mapping, resulting in a complete 3D model. This simple method effectively transforms the 3D generation problem into a 2D image generation task, maximizing the utilization of knowledge in pretrained diffusion models. Furthermore, we demonstrate that our Kiss3DGen model is compatible with various diffusion model techniques, enabling advanced features such as 3D editing, mesh and texture enhancement, etc. Through extensive experiments, we demonstrate the effectiveness of our approach, showcasing its ability to produce high-quality 3D models efficiently Project page: https://ltt-0.github.io/Kiss3dgen.github.io.

Despite the growing integration of deep models into mobile terminals, the accuracy of these models declines significantly due to various deployment interferences. Test-time adaptation (TTA) has emerged to improve the performance of deep models by adapting them to unlabeled target data online. Yet, the significant memory cost, particularly in resource-constrained terminals, impedes the effective deployment of most backward-propagation-based TTA methods. To tackle memory constraints, we introduce Surgeon, a method that substantially reduces memory cost while preserving comparable accuracy improvements during fully test-time adaptation (FTTA) without relying on specific network architectures or modifications to the original training procedure. Specifically, we propose a novel dynamic activation sparsity strategy that directly prunes activations at layer-specific dynamic ratios during adaptation, allowing for flexible control of learning ability and memory cost in a data-sensitive manner. Among this, two metrics, Gradient Importance and Layer Activation Memory, are considered to determine the layer-wise pruning ratios, reflecting accuracy contribution and memory efficiency, respectively. Experimentally, our method surpasses the baselines by not only reducing memory usage but also achieving superior accuracy, delivering SOTA performance across diverse datasets, architectures, and tasks.

Text-to-video generative models have made significant strides, enabling diverse applications in entertainment, advertising, and education. However, generating RGBA video, which includes alpha channels for transparency, remains a challenge due to limited datasets and the difficulty of adapting existing models. Alpha channels are crucial for visual effects (VFX), allowing transparent elements like smoke and reflections to blend seamlessly into scenes. We introduce TransPixeler, a method to extend pretrained video models for RGBA generation while retaining the original RGB capabilities. TransPixeler leverages a diffusion transformer (DiT) architecture, incorporating alpha-specific tokens and using LoRA-based fine-tuning to jointly generate RGB and alpha channels with high consistency. By optimizing attention mechanisms, TransPixeler preserves the strengths of the original RGB model and achieves strong alignment between RGB and alpha channels despite limited training data. Our approach effectively generates diverse and consistent RGBA videos, advancing the possibilities for VFX and interactive content creation. The code is available at https://wileewang.github.io/TransPixeler/.

Video-to-audio synthesis, which generates synchronized audio for visual content, critically enhances viewer immersion and narrative coherence in film and interactive media.However, video-to-audio dubbing for long-form content remains an unsolved challenge due to dynamic semantic shifts, audio diversity and the absence of dedicated datasets.While existing methods excel in short videos, they falter in long scenarios (e.g., movies) due to fragmented synthesis and inadequate cross-scene consistency.We propose LVAS-Agent, a multi-agent framework that offers a coordinated, multi-component approach to long-video audio generation.Our approach decomposes long-video synthesis into four steps including scene segmentation, script generation, audio design and audio synthesis.To enable systematic evaluation, we introduce LVAS-Bench, the first benchmark with 207 professionally curated long videos spanning diverse scenarios.Experiments show that our method outperforms state-of-the-art V2A models in overall audio synthesis quality.

We propose PRM, a novel photometric stereo based large reconstruction model to reconstruct high-quality meshes with fine-grained local details. Unlike previous large reconstruction models that prepare images under fixed and simple lighting as both input and supervision, PRM renders photometric stereo images by varying materials and lighting for the purposes, which not only improves the precise local details by providing rich photometric cues but also increases the model robustness to variations in the appearance of input images. To offer enhanced flexibility of images rendering, we incorporate a real-time physically-based rendering (PBR) method and mesh rasterization for online images rendering. Moreover, in employing an explicit mesh as our 3D representation, PRM ensures the application of differentiable PBR, which supports the utilization of multiple photometric supervisions and better models the specular color for high-quality geometry optimization. Our PRM leverages photometric stereo images to achieve high-quality reconstructions with fine-grained local details, even amidst sophisticated image appearances. Extensive experiments demonstrate that PRM significantly outperforms other models.

Recent advancements in generative models have enabled 3D urban scene generation from satellite imagery, unlocking promising applications in gaming, digital twins, and beyond. However, most existing methods rely heavily on neural rendering techniques, which hinder their ability to produce detailed 3D structures on a broader scale, largely due to the inherent structural ambiguity derived from relatively limited 2D observations. To address this challenge, we propose Sat2City, a novel framework that synergizes the representational capacity of sparse voxel grids with latent diffusion models, tailored specifically for our novel 3D city dataset. Our approach is enabled by three key components: (1) A cascaded latent diffusion framework that progressively recovers 3D city structures from satellite imagery, (2) a Re-Hash operation at its Variational Autoencoder (VAE) bottleneck to compute multi-scale feature grids for stable appearance optimization and (3) an inverse sampling strategy enabling implicit supervision for smooth appearance transitioning.To overcome the challenge of collecting real-world city-scale 3D models with high-quality geometry and appearance, we introduce a dataset of synthesized large-scale 3D cities paired with satellite-view height maps. Validated on this dataset, our framework generates detailed 3D structures from a single satellite image, achieving superior fidelity compared to existing city generation models.

In this work, we introduce FlexGen, a flexible framework designed to generate controllable and consistent multi-view images, conditioned on a single-view image, or a text prompt, or both. FlexGen tackles the challenges of controllable multi-view synthesis through additional conditioning on 3D-aware text annotations. We utilize the strong reasoning capabilities of GPT-4V to generate 3D-aware text annotations. By analyzing four orthogonal views of an object arranged as tiled multi-view images, GPT-4V can produce text annotations that include 3D-aware information with spatial relationship. By integrating the control signal with proposed adaptive dual-control module, our model can generate multi-view images that correspond to the specified text. FlexGen supports multiple controllable capabilities, allowing users to modify text prompts to generate reasonable and corresponding unseen parts. Additionally, users can influence attributes such as appearance and material properties, including metallic and roughness. Extensive experiments demonstrate that our approach offers enhanced multiple controllability, marking a significant advancement over existing multi-view diffusion models. This work has substantial implications for fields requiring rapid and flexible 3D content creation, including game development, animation, and virtual reality. Project page: https://xxu068.github.io/flexgen.github.io/.

Estimating physical properties for visual data is a crucial task in computer vision, graphics, and robotics, underpinning applications such as augmented reality, physical simulation, and robotic grasping. However, this area remains under-explored due to the inherent ambiguities in physical property estimation. To address these challenges, we introduce GaussianProperty, a training-free framework that assigns physical properties of materials to 3D Gaussians. Specifically, we integrate the segmentation capability of SAM with the recognition capability of GPT-4V(ision) to formulate a global-local physical property reasoning module for 2D images. Then we project the physical properties from multi-view 2D images to 3D Gaussians using a voting strategy. We demonstrate that 3D Gaussians with physical property annotations enable applications in physics-based dynamic simulation and robotic grasping. For physics-based dynamic simulation, we leverage the Material Point Method (MPM) for realistic dynamic simulation. For robot grasping, we develop a grasping force prediction strategy that estimates a safe force range required for object grasping based on the estimated physical properties. Extensive experiments on material segmentation, physics-based dynamic simulation, and robotic grasping validate the effectiveness of our proposed method, highlighting its crucial role in understanding physical properties from visual data. Online demo, code, more cases and annotated datasets are available on \href{https://Gaussian-Property.github.io}{this https URL}.

Leveraging the visual priors of pre-trained text-to-image diffusion models offers a promising solution to enhance zero-shot generalization in dense prediction tasks. However, existing methods often uncritically use the original diffusion formulation, which may not be optimal due to the fundamental differences between dense prediction and image generation. In this paper, we provide a systemic analysis of the diffusion formulation for the dense prediction, focusing on both quality and efficiency. And we find that the original parameterization type for image generation, which learns to predict noise, is harmful for dense prediction; the multi-step noising/denoising diffusion process is also unnecessary and challenging to optimize. Based on these insights, we introduce $\textbf{Lotus}$, a diffusion-based visual foundation model with a simple yet effective adaptation protocol for dense prediction. Specifically, Lotus is trained to directly predict annotations instead of noise, thereby avoiding harmful variance. We also reformulate the diffusion process into a single-step procedure, simplifying optimization and significantly boosting inference speed. Additionally, we introduce a novel tuning strategy called detail preserver, which achieves more accurate and fine-grained predictions. Without scaling up the training data or model capacity, Lotus achieves SoTA performance in zero-shot depth and normal estimation across various datasets. It also enhances efficiency, being significantly faster than most existing diffusion-based methods. Lotus' superior quality and efficiency also enable a wide range of practical applications, such as joint estimation, single/multi-view 3D reconstruction, etc.

In the realm of image generation, creating customized images from visual prompt with additional textual instruction emerges as a promising endeavor. However, existing methods, both tuning-based and tuning-free, struggle with interpreting the subject-essential attributes from the visual prompt. This leads to subject-irrelevant attributes infiltrating the generation process, ultimately compromising the personalization quality in both editability and ID preservation. In this paper, we present $\textbf{DisEnvisioner}$, a novel approach for effectively extracting and enriching the subject-essential features while filtering out -irrelevant information, enabling exceptional customization performance, in a $\textbf{tuning-free}$ manner and using only $\textbf{a single image}$. Specifically, the feature of the subject and other irrelevant components are effectively separated into distinctive visual tokens, enabling a much more accurate customization. Aiming to further improving the ID consistency, we enrich the disentangled features, sculpting them into a more granular representation. Experiments demonstrate the superiority of our approach over existing methods in instruction response (editability), ID consistency, inference speed, and the overall image quality, highlighting the effectiveness and efficiency of DisEnvisioner.

Multi-objective test-time alignment aims to adapt large language models (LLMs) to diverse multi-dimensional user preferences during inference while keeping LLMs frozen. Recently, GenARM (Xu et al., 2025) first independently trains Autoregressive Reward Models (ARMs) for each preference dimension without awareness of each other, then combines their outputs based on user-specific preference vectors during inference to achieve multi-objective test-time alignment, leading to two key limitations: the need for multiple ARMs increases the inference cost, and the separate training of ARMs causes the misalignment between the guided generation and the user preferences. To address these issues, we propose Preference-aware ARM (PARM), a single unified ARM trained across all preference dimensions. PARM uses our proposed Preference-Aware Bilinear Low-Rank Adaptation (PBLoRA), which employs a bilinear form to condition the ARM on preference vectors, enabling it to achieve precise control over preference trade-offs during inference. Experiments demonstrate that PARM reduces inference costs and achieves better alignment with preference vectors compared with existing methods. Additionally, PARM enables weak-to-strong guidance, allowing a smaller PARM to guide a larger frozen LLM without expensive training, making multi-objective alignment accessible with limited computing resources. The code is available at https://github.com/Baijiong-Lin/PARM.

Large Language Models (LLMs) have made substantial advancements in the field of robotic and autonomous driving. This study presents the first Occupancy-based Large Language Model (Occ-LLM), which represents a pioneering effort to integrate LLMs with an important representation. To effectively encode occupancy as input for the LLM and address the category imbalances associated with occupancy, we propose Motion Separation Variational Autoencoder (MS-VAE). This innovative approach utilizes prior knowledge to distinguish dynamic objects from static scenes before inputting them into a tailored Variational Autoencoder (VAE). This separation enhances the model's capacity to concentrate on dynamic trajectories while effectively reconstructing static scenes. The efficacy of Occ-LLM has been validated across key tasks, including 4D occupancy forecasting, self-ego planning, and occupancybased scene question answering. Comprehensive evaluations demonstrate that Occ-LLM significantly surpasses existing state-of-the-art methodologies, achieving gains of about 6% in Intersection over Union (IoU) and 4% in mean Intersection over Union (mIoU) for the task of 4D occupancy forecasting. These findings highlight the transformative potential of Occ-LLM in reshaping current paradigms within robotic and autonomous driving.

3D point cloud segmentation aims to assign semantic labels to individual points in a scene for fine-grained spatial understanding. Existing methods typically adopt data augmentation to alleviate the burden of large-scale annotation. However, most augmentation strategies only focus on local transformations or semantic recomposition, lacking the consideration of global structural dependencies within scenes. To address this limitation, we propose a graph-guided data augmentation framework with dual-level constraints for realistic 3D scene synthesis. Our method learns object relationship statistics from real-world data to construct guiding graphs for scene generation. Local-level constraints enforce geometric plausibility and semantic consistency between objects, while global-level constraints maintain the topological structure of the scene by aligning the generated layout with the guiding graph. Extensive experiments on indoor and outdoor datasets demonstrate that our framework generates diverse and high-quality augmented scenes, leading to consistent improvements in point cloud segmentation performance across various models. Code is available at: https://github.com/alexander7xu/DualLevelAug

Recent advancements in low-light video enhancement (LLVE) have increasingly leveraged both RGB and event cameras to improve video quality under challenging conditions. However, existing approaches share two key drawbacks. First, they are tuned for steady low-light scenes, so their performance drops when illumination varies. Second, they assume every sensing modality is always available, while real systems may lose or corrupt one of them. These limitations make the methods brittle in dynamic, real-world settings. In this paper, we propose EVDiffuser, a novel framework for consistent LLVE that integrates RGB and event data through a modality-adaptive diffusion pipeline. By harnessing the powerful priors of video diffusion models, EVDiffuser enables consistent video enhancement and generalization to diverse scenarios under varying illumination, where RGB or events may even be absent. Specifically, we first design a modality-agnostic conditioning mechanism based on a diffusion pipeline by treating the two modalities as optional conditions, which is fine-tuned using augmented and integrated datasets. Furthermore, we introduce a modality-adaptive guidance rescaling that dynamically adjusts the contribution of each modality according to sensor-specific characteristics. Additionally, we establish a benchmark that accounts for varying illumination and diverse real-world scenarios, facilitating future research on consistent event-guided LLVE. Our experiments demonstrate state-of-the-art performance across challenging scenarios (i.e., varying illumination) and sensor-based settings (e.g., event-only, RGB-only), highlighting the generalization of our framework.

With the rapid advancement of generative models, general-purpose generation has gained increasing attention as a promising approach to unify diverse tasks across modalities within a single system. Despite this progress, existing open-source frameworks often remain fragile and struggle to support complex real-world applications due to the lack of structured workflow planning and execution-level feedback. To address these limitations, we present ComfyMind, a collaborative AI system designed to enable robust and scalable general-purpose generation, built on the ComfyUI platform. ComfyMind introduces two core innovations: Semantic Workflow Interface (SWI) that abstracts low-level node graphs into callable functional modules described in natural language, enabling high-level composition and reducing structural errors; Search Tree Planning mechanism with localized feedback execution, which models generation as a hierarchical decision process and allows adaptive correction at each stage. Together, these components improve the stability and flexibility of complex generative workflows. We evaluate ComfyMind on three public benchmarks: ComfyBench, GenEval, and Reason-Edit, which span generation, editing, and reasoning tasks. Results show that ComfyMind consistently outperforms existing open-source baselines and achieves performance comparable to GPT-Image-1. ComfyMind paves a promising path for the development of open-source general-purpose generative AI systems.

Large reconstruction model has remarkable progress, which can directly predict 3D or 4D representations for unseen scenes and objects. However, current work has not systematically explored the potential of large reconstruction models in the field of autonomous driving. To achieve this, we introduce the Large 4D Gaussian Reconstruction Model (DrivingRecon). With an elaborate and simple framework design, it not only ensures efficient and high-quality reconstruction, but also provides potential for downstream tasks. There are two core contributions: firstly, the Prune and Dilate Block (PD-Block) is proposed to prune redundant and overlapping Gaussian points and dilate Gaussian points for complex objects. Then, dynamic and static decoupling is tailored to better learn the temporary-consistent geometry across different time. Experimental results demonstrate that DrivingRecon significantly improves scene reconstruction quality compared to existing methods. Furthermore, we explore applications of DrivingRecon in model pre-training, vehicle type adaptation, and scene editing. Our code will be available.

Robust and unobtrusive in-vehicle physiological monitoring is crucial for ensuring driving safety and user experience. While remote physiological measurement (RPM) offers a promising non-invasive solution, its translation to real-world driving scenarios is critically constrained by the scarcity of comprehensive datasets. Existing resources are often limited in scale, modality diversity, the breadth of biometric annotations, and the range of captured conditions, thereby omitting inherent real-world challenges in driving. Here, we present PhysDrive, the first large-scale multimodal dataset for contactless in-vehicle physiological sensing with dedicated consideration of various modality settings and driving factors. PhysDrive collects data from 48 drivers, including synchronized RGB, near-infrared camera, and raw mmWave radar data, accompanied by six synchronized ground truths (ECG, BVP, Respiration, HR, RR, and SpO2). It covers a wide spectrum of naturalistic driving conditions, including driver motions, dynamic natural light, vehicle types, and road conditions. We extensively evaluate both signal‑processing and deep‑learning methods on PhysDrive, establishing a comprehensive benchmark across all modalities, and release full open‑source code with compatibility for mainstream public toolboxes. We envision PhysDrive will serve as a foundational resource and accelerate research on multimodal driver monitoring and smart‑cockpit systems.

MOTIVATION: Structure-based protein design is crucial for designing proteins with novel structures and functions, which aims to generate sequences that fold into desired structures. Current deep learning-based methods primarily focus on training and evaluating models using sequence recovery-based metrics. However, this approach overlooks the inherent ambiguity in the relationship between protein sequences and structures. Relying solely on sequence recovery as a training objective limits the models' ability to produce diverse sequences that maintain similar structures. These limitations become more pronounced when dealing with remote homologous proteins, which share functional and structural similarities despite low-sequence identity. RESULTS: Here, we present DivPro, a model that learns to design diverse sequences that can fold into similar structures. To improve sequence diversity, instead of learning a single fixed sequence representation for an input structure as in existing methods, DivPro learns a probabilistic sequence space from which diverse sequences could be sampled. We leverage the recent advancements in in silico protein structure prediction. By incorporating structure prediction results as training guidance, DivPro ensures that sequences sampled from this learned space reliably fold into the target structure. We conducted extensive experiments on three sequence design benchmarks and evaluated the structures of designed sequences using structure prediction models including AlphaFold2. Results show that DivPro can maintain high structure recovery while significantly improving the sequence diversity. AVAILABILITY AND IMPLEMENTATION: The source code and datasets are available at https://github.com/veghen/DivPro.

Abstract Recent large reconstruction models have made notable progress in generating high‐quality 3D objects from single images. However, current reconstruction methods often rely on explicit camera pose estimation or fixed viewpoints, restricting their flexibility and practical applicability. We reformulate 3D reconstruction as image‐to‐image translation and introduce the Relative Coordinate Map (RCM), which aligns multiple unposed images to a “main” view without pose estimation. While RCM simplifies the process, its lack of global 3D supervision can yield noisy outputs. To address this, we propose Relative Coordinate Gaussians (RCG) as an extension to RCM, which treats each pixel's coordinates as a Gaussian center and employs differentiable rasterization for consistent geometry and pose recovery. Our LucidFusion framework handles an arbitrary number of unposed inputs, producing robust 3D reconstructions within seconds and paving the way for more flexible, pose‐free 3D pipelines.

Multi-task dense scene understanding, which trains a model for multiple dense prediction tasks, has a wide range of application scenarios. Capturing long-range dependency and enhancing cross-task interactions are crucial to multi-task dense prediction. In this paper, we propose MTMamba++, a novel architecture for multi-task scene understanding featuring with a Mamba-based decoder. It contains two types of core blocks: self-task Mamba (STM) block and cross-task Mamba (CTM) block. STM handles long-range dependency by leveraging state-space models, while CTM explicitly models task interactions to facilitate information exchange across tasks. We design two types of CTM block, namely F-CTM and S-CTM, to enhance cross-task interaction from feature and semantic perspectives, respectively. Extensive experiments on NYUDv2, PASCAL-Context, and Cityscapes datasets demonstrate the superior performance of MTMamba++ over CNN-based, Transformer-based, and diffusion-based methods while maintaining high computational efficiency.

Remote photoplethysmography (rPPG) has been widely applied to measure heart rate from face videos. To increase the generalizability of the algorithms, domain generalization (DG) attracted increasing attention in rPPG. However, when rPPG is extended to simultaneously measure more vital signs (e.g., respiration and blood oxygen saturation), achieving generalizability brings new challenges. Although partial features shared among different physiological signals can benefit multi-task learning, the sparse and imbalanced target label space brings the seesaw effect over task-specific feature learning. To resolve this problem, we designed an end-to-end Mixture of Low-rank Experts for multi-task remote Physiological measurement (PhysMLE), which is based on multiple low-rank experts with a novel router mechanism, thereby enabling the model to adeptly handle both specifications and correlations within tasks. Additionally, we introduced prior knowledge from physiology among tasks to overcome the imbalance of label space under real-world multi-task physiological measurement. For fair and comprehensive evaluations, this paper proposed a large-scale multi-task generalization benchmark, named Multi-Source Synsemantic Domain Generalization (MSSDG) protocol. Extensive experiments with MSSDG and intra-dataset have shown the effectiveness and efficiency of PhysMLE. In addition, a new dataset was collected and made publicly available to meet the needs of the MSSDG. The code and data are available at https://github.com/WJULYW/PhysMLE.

Remote photoplethysmography (rPPG) is a promising technology for capturing physiological signals from facial videos, with potential applications in medical health, affective computing, and biometric recognition. The demand for rPPG tasks has evolved from achieving high performance in intra-dataset testing to excelling in cross-dataset testing (i.e., domain generalization). However, most existing methods have overlooked the incorporation of prior knowledge specific to rPPG, leading to limited generalization capabilities. In this paper, we propose a novel framework that effectively integrates both explicit and implicit prior knowledge into the rPPG task. Specifically, we conduct a systematic analysis of noise sources (e.g., variations in cameras, lighting conditions, skin types, and motion) across different domains and embed this prior knowledge into the network design. Furthermore, we employ a two-branch network to disentangle physiological feature distributions from noise through implicit label correlation. Extensive experiments demonstrate that the proposed method not only surpasses state-of-the-art approaches in RGB cross-dataset evaluation but also exhibits strong generalization from RGB datasets to NIR datasets. The code is publicly available at https://github.com/keke-nice/Greip.

In this paper, we study a new problem, Film Removal (FR), which attempts to remove the interference of wrinkled transparent films and reconstruct the original information under films for industrial recognition systems. We first physically model the imaging of industrial materials covered by the film. Considering the specular highlight from the film can be effectively recorded by the polarized camera, we build a practical dataset with polarization information containing paired data with and without transparent film. We aim to remove interference from the film (specular highlights and other degradations) with an end-to-end framework. To locate the specular highlight, we use an angle estimation network to optimize the polarization angle with the minimized specular highlight. The image with minimized specular highlight is set as a prior for supporting the reconstruction network. Based on the prior and the polarized images, the reconstruction network can decouple all degradations from the film. Extensive experiments show that our framework achieves SOTA performance in both image reconstruction and industrial downstream tasks. Our code will be released at https://github.com/jqtangust/FilmRemoval.

The recent advancements in text-to-3D generation mark a significant milestone in generative models, unlocking new possibilities for creating imaginative 3D assets across var-ious real-world scenarios. While recent advancements in text-to-3D generation have shown promise, they often fall short in rendering detailed and high-quality 3D models. This problem is especially prevalent as many methods base themselves on Score Distillation Sampling (SDS). This paper identifies a notable deficiency in SDS, that it brings inconsistent and low-quality updating direction for the 3D model, causing the over-smoothing effect. To address this, we propose a novel approach called Interval Score Matching (ISM). ISM employs deterministic diffusing trajectories and utilizes interval-based score matching to counteract over-smoothing. Furthermore, we incorporate 3D Gaussian Splatting into our text-to-3D generation pipeline. Extensive experiments show that our model largely outperforms the state-of-the-art in quality and training efficiency. Our code is available at: EnVision-Research/LucidDreamer

This paper focuses on the high computational complexity in Large Language Models (LLMs), a significant challenge in both natural language processing (NLP) and multi-modal tasks. We propose Low-Rank Approximation for Sparse Attention (LoRA -Sparse), an innovative approach that strategically reduces this complexity. LoRA -Sparse introduces low-rank linear projection layers for sparse attention approximation. It utilizes an order-mimic training methodology, which is crucial for efficiently approximating the self-attention mechanism in LLMs. We empirically show that sparse attention not only reduces computational demands, but also enhances model performance in both NLP and multi-modal tasks. This surprisingly shows that redundant attention in LLMs might be non-beneficial. We extensively validate LoRA -Sparse through rigorous empirical studies in both (NLP) and multi-modal tasks, demonstrating its effectiveness and general applicability. Based on LLaMA and LLaVA models, our methods can reduce more than half of the self-attention computation with even better performance than full-attention baselines.

Though graph neural networks (GNNs) have been used in SAT solution prediction, for a subset of symmetric SAT problems, we unveil that the current GNN-based end-to-end SAT solvers are bound to yield incorrect outcomes as they are unable to break symmetry in variable assignments. In response, we introduce AsymSAT, a new GNN architecture coupled where a recurrent neural network is (RNN) to produce asymmetric models. Moreover, we bring up a method to integrate machine-learning-based SAT assignment prediction with classic SAT solvers and demonstrate its performance on non-trivial SAT instances including logic equivalence checking and cryptographic analysis problems with as much as 75.45% time saving.

Contrastive Learning (CL) has attracted enormous attention due to its remarkable capability in unsupervised representation learning. However, recent works have revealed the vulnerability of CL to backdoor attacks: the feature extractor could be misled to embed backdoored data close to an attack target class, thus fooling the downstream predictor to misclassify it as the target. Existing attacks usually adopt a fixed trigger pattern and poison the training set with trigger-injected data, hoping for the feature extractor to learn the association between trigger and target class. However, we find that such fixed trigger design fails to effectively associate trigger-injected data with target class in the embedding space due to special CL mechanisms, leading to a limited attack success rate (ASR). This phenomenon motivates us to find a better backdoor trigger design tailored for CL framework. In this paper, we propose a bi-level optimization approach to achieve this goal, where the inner optimization simulates the CL dynamics of a surrogate victim, and the outer optimization enforces the backdoor trigger to stay close to the target throughout the surrogate CL procedure. Extensive experiments show that our attack can achieve a higher attack success rate (e.g., 99\% ASR on ImageNet-100) with a very low poisoning rate (1\%). Besides, our attack can effectively evade existing state-of-the-art defenses.

Denoising Diffusion Probabilistic Models (DDPMs) have garnered popularity for data generation across various domains. However, a significant bottleneck is the necessity for whole-network computation during every step of the generative process, leading to high computational overheads. This paper presents a novel framework, Denoising Diffusion Step-aware Models (DDSM), to address this challenge. Unlike conventional approaches, DDSM employs a spectrum of neural networks whose sizes are adapted according to the importance of each generative step, as determined through evolutionary search. This step-wise network variation effectively circumvents redundant computational efforts, particularly in less critical steps, thereby enhancing the efficiency of the diffusion model. Furthermore, the step-aware design can be seamlessly integrated with other efficiency-geared diffusion models such as DDIMs and latent diffusion, thus broadening the scope of computational savings. Empirical evaluations demonstrate that DDSM achieves computational savings of 49% for CIFAR-10, 61% for CelebA-HQ, 59% for LSUN-bedroom, 71% for AFHQ, and 76% for ImageNet, all without compromising the generation quality. Our code and models are available at https://github.com/EnVision-Research/DDSM.

Learning surfaces from neural radiance field (NeRF) became a rising topic in Multi-View Stereo (MVS). Recent Signed Distance Function (SDF)-based methods demonstrated their ability to reconstruct exact 3D shapes of Lambertian scenes. However, their results on reflective scenes are unsatisfactory due to the entanglement of specular radiance and complicated geometry. To address the challenges, we propose a Gaussian-based representation of normals in SDF fields. Supervised by polarization priors, this representation guides the learning of geometry behind the specular reflection and capture more details than existing methods. Moreover, we propose a reweighting strategy in optimization process to alleviate the noise issue of polarization priors. To validate the effectiveness of our design, we capture polarimetric information and ground truth meshes in additional reflective scenes with various geometry. We also evaluated our framework on PANDORA dataset. Both qualitative and quantitative comparisons prove our method outperforms existing neural 3D reconstruction methods in reflective scenes by a large margin.

Diffusion Probabilistic Models (DPMs) show significant potential in image generation, yet their performance hinges on having access to large datasets. Previous works, like Generative Adversarial Networks (GANs), have tackled the limited data problem by transferring pre-trained models learned with sufficient data. However, those methods are hard to be utilized in DPMs since the distinct differences between DPM-based and GAN-based methods, showing in the unique iterative denoising process integral and the need for many timesteps with no-targeted noise in DPMs. In this paper, we propose a novel DPMs-based transfer learning method, ANT, to address the limited data problem. It includes two strategies: similarity-guided training, which boosts transfer with a classifier, and adversarial noise selection which adaptively chooses targeted noise based on the input image. Extensive experiments in the context of few-shot image generation tasks demonstrate that our method is not only efficient but also excels in terms of image quality and diversity when compared to existing GAN-based and DDPM-based methods.

We explore Bird’s-Eye View (BEV) generation, converting a BEV map into its corresponding multi-view street images. Valued for its unified spatial representation aiding multi-sensor fusion, BEV is pivotal for various autonomous driving applications. Creating accurate street-view images from BEV maps is essential for portraying complex traffic scenarios and enhancing driving algorithms. Concurrently, diffusion-based conditional image generation models have demonstrated remarkable outcomes, adept at producing diverse, high-quality, and condition-aligned results. Nonetheless, the training of these models demands substantial data and computational resources. Hence, exploring methods to fine-tune these advanced models, like Stable Diffusion, for specific conditional generation tasks emerges as a promising avenue. In this paper, we introduce a practical framework for generating images from a BEV layout. Our approach comprises two main components: the Neural View Transformation and the Street Image Generation. The Neural View Transformation phase converts the BEV map into aligned multi-view semantic segmentation maps by learning the shape correspondence between the BEV and perspective views. Subsequently, the Street Image Generation phase utilizes these segmentations as a condition to guide a fine-tuned latent diffusion model. This finetuning process ensures both view and style consistency. Our model leverages the generative capacity of large pretrained diffusion models within traffic contexts, effectively yielding diverse and condition-coherent street view images.

Deep neural networks (DNNs) have been proven extremely susceptible to adversarial examples, which raises special safety-critical concerns for DNN-based autonomous driving stacks (i.e., 3D object detection). Although there are extensive works on image-level attacks, most are restricted to 2D pixel spaces, and such attacks are not always physically realistic in our 3D world. Here we present Adv3D, the first exploration of modeling adversarial examples as Neural Radiance Fields (NeRFs) in driving scenarios. Advances in NeRF provide photorealistic appearances and 3D accurate generation, yielding a more realistic and realizable adversarial example. We train our adversarial NeRF by minimizing the surrounding objects’ confidence predicted by 3D detectors on the training set. Then we evaluate Adv3D on the unseen validation set and show that it can cause a large performance reduction when rendering NeRF in any sampled pose. To enhance physical effectiveness, we propose primitive-aware sampling and semantic-guided regularization that enable 3D patch attacks with camouflage adversarial texture. Experimental results demonstrate that our method surpasses the mesh baseline and generalizes well to different poses, scenes, and 3D detectors. Finally, we provide a defense method to our attacks that improves both the robustness and clean performance of 3D detectors.

Remote photoplethysmography (rPPG) is a promising technique for non-contact physiological signal measurement. It has great potential applications in human health monitoring and emotion analysis. However, existing methods for the rPPG task ignore the long-tail phenomenon of physiological signal data, especially on multi-domain joint training. In addition, we find that the long-tail problem of the physiological label (phys-label) exists in different datasets, and the long-tail problem of some domain exists under the same phys-label. To tackle these problems, we propose a hierarchical balanced framework, to mitigate the bias caused by domain and phys-label imbalance. Specifically, we propose anti-spurious domain center learning tailored to learning domain-balanced embeddings space. Then, we adopt compact-aware continuity regularization to estimate phys-label-wise imbalances and construct continuity between embeddings. Extensive experiments demonstrate that our method outperforms the state-of-the-art in cross-dataset and intra-dataset settings. Our code is available at https://github.com/pywin/HiBa.

Video Anomaly Detection (VAD) systems can autonomously monitor and identify disturbances, reducing the need for manual labor and associated costs. However, current VAD systems are often limited by their superficial semantic understanding of scenes and minimal user interaction. Additionally, the prevalent data scarcity in existing datasets restricts their applicability in open-world scenarios. In this paper, we introduce HAWK, a novel framework that leverages interactive large Visual Language Models (VLM) to interpret video anomalies precisely. Recognizing the difference in motion information between abnormal and normal videos, HAWK explicitly integrates motion modality to enhance anomaly identification. To reinforce motion attention, we construct an auxiliary consistency loss within the motion and video space, guiding the video branch to focus on the motion modality. Moreover, to improve the interpretation of motion-to-language, we establish a clear supervisory relationship between motion and its linguistic representation. Furthermore, we have annotated over 8,000 anomaly videos with language descriptions, enabling effective training across diverse open-world scenarios, and also created 8,000 question-answering pairs for users' open-world questions. The final results demonstrate that HAWK achieves SOTA performance, surpassing existing baselines in both video description generation and question-answering. Our codes/dataset/demo will be released at https://github.com/jqtangust/hawk.

The utilization of remote photoplethysmography (rPPG) technology has gained attention in recent years due to its ability to extract blood volume pulse (BVP) from facial videos, making it accessible for various applications such as health monitoring and emotional analysis. However, the BVP signal is susceptible to complex environmental changes or individual differences, causing existing methods to struggle in generalizing for unseen domains. This article addresses the domain shift problem in rPPG measurement and shows that most domain generalization methods fail to work well in this problem due to ambiguous instance-specific differences. To address this, the article proposes a novel approach called Hierarchical Style-aware Representation Disentangling (HSRD). HSRD improves generalization capacity by separating domain-invariant and instance-specific feature space during training, which increases the robustness of out-of-distribution samples during inference. This work presents state-of-the-art performance against several methods in both cross and intra-dataset settings.

Remote photoplethysmography (rPPG) is a contactless technique that facilitates the measurement of physiological signals and cardiac activities through facial video recordings. This approach holds tremendous potential for various applications. However, existing rPPG methods often did not account for different types of occlusions that commonly occur in real-world scenarios, such as temporary movement or actions of humans in videos or dust on camera. The failure to address these occlusions can compromise the accuracy of rPPG algorithms. To address this issue, we proposed a novel Condiff-rPPG to improve the robustness of rPPG measurement facing various occlusions. First, we compressed the damaged face video into a spatio-temporal representation with several types of masks. Second, the diffusion model was designed to recover the missing information with observed values as a condition. Moreover, a novel low-rank decomposition regularization was proposed to eliminate background noise and maximize informative features. ConDiff-rPPG ensured consistency in optimization goals during the training process. Through extensive experiments, including intra- and cross-dataset evaluations, as well as ablation tests, we demonstrated the robustness and generalization ability of our proposed model.

Remote photoplethysmography (rPPG) is a non-invasive technique that aims to capture subtle variations in facial pixels caused by changes in blood volume resulting from cardiac activities. Most existing unsupervised methods for rPPG tasks focus on the contrastive learning between samples while neglecting the inherent self-similarity prior in physiological signals. In this paper, we propose a Self-Similarity Prior Distillation (SSPD) framework for unsupervised rPPG estimation, which capitalizes on the intrinsic temporal self-similarity of cardiac activities. Specifically, we first introduce a physical-prior embedded augmentation technique to mitigate the effect of various types of noise. Then, we tailor a self-similarity-aware network to disentangle more reliable self-similar physiological features. Finally, we develop a hierarchical self-distillation paradigm for self-similarity-aware learning and rPPG signal decoupling. Comprehensive experiments demonstrate that the unsupervised SSPD framework achieves comparable or even superior performance compared to the state-of-the-art supervised methods. Meanwhile, SSPD has the lowest inference time and computation cost among end-to-end models.

Remote photoplethysmography (rPPG) technology has drawn increasing attention in recent years. It can extract Blood Volume Pulse (BVP) from facial videos, making many applications like health monitoring and emotional analysis more accessible. However, as the BVP signal is easily affected by environmental changes, existing methods struggle to generalize well for unseen domains. In this paper, we systematically address the domain shift problem in the rPPG measurement task. We show that most domain generalization methods do not work well in this problem, as domain labels are ambiguous in complicated environmental changes. In light of this, we propose a domain-label-free approach called NEuron STructure modeling (NEST). NEST improves the generalization capacity by maximizing the coverage of feature space during training, which reduces the chance for under-optimized feature activation during inference. Besides, NEST can also enrich and enhance domain invariant features across multi-domain. We create and benchmark a large-scale domain generalization protocol for the rPPG measurement task. Extensive experiments show that our approach outperforms the state-of-the-art methods on both cross-dataset and intra-dataset settings. The codes are available at https://github.com/LuPaoPao/NEST.

This work explores the use of 3D generative models to synthesize training data for 3D vision tasks. The key requirements of the generative models are that the generated data should be photorealistic to match the real-world scenarios, and the corresponding 3D attributes should be aligned with given sampling labels. However, we find that the recent NeRF-based 3D GANs hardly meet the above requirements due to their designed generation pipeline and the lack of explicit 3D supervision. In this work, we propose Lift3D, an inverted 2D-to-3D generation framework to achieve the data generation objectives. Lift3D has several merits compared to prior methods: (1) Unlike previous 3D GANs that the output resolution is fixed after training, Lift3D can generalize to any camera intrinsic with higher resolution and photorealistic output. (2) By lifting well-disentangled 2D GAN to 3D object NeRF, Lift3D provides explicit 3D information of generated objects, thus offering accurate 3D annotations for downstream tasks. We evaluate the effectiveness of our framework by augmenting autonomous driving datasets. Experimental results demonstrate that our data generation framework can effectively improve the performance of 3D object detectors. Code: len-li.github.io/lift3d-web

Contemporary image rescaling aims at embedding a high-resolution (HR) image into a low-resolution (LR) thumbnail image that contains embedded information for HR image reconstruction. Unlike traditional image super-resolution, this enables high-fidelity HR image restoration faithful to the original one, given the embedded information in the LR thumbnail. However, state-of-the-art image rescaling methods do not optimize the LR image file size for efficient sharing and fall short of real-time performance for ultra-high-resolution (e.g., 6K) image reconstruction. To address these two challenges, we propose a novel frame-work (HyperThumbnail) for real-time 6K rate-distortion-aware image rescaling. Our framework first embeds an HR image into a JPEG LR thumbnail by an encoder with our proposed quantization prediction module, which minimizes the file size of the embedding LR JPEG thumbnail while maximizing HR reconstruction quality. Then, an efficient frequency-aware decoder reconstructs a high-fidelity HR image from the LR one in real time. Extensive experiments demonstrate that our framework outperforms previous image rescaling baselines in rate-distortion performance and can perform 6K image reconstruction in real time.

Due to limited camera capacities, digital images usually have a narrower dynamic illumination range than real-world scene radiance. To resolve this problem, High Dynamic Range (HDR) reconstruction is proposed to recover the dynamic range to better represent real-world scenes. However, due to different physical imaging parameters, the tone-mapping functions between images and real radiance are highly diverse, which makes HDR reconstruction extremely challenging. Existing solutions can not explicitly clarify a corresponding relationship between the tone-mapping function and the generated HDR image, but this relationship is vital when guiding the reconstruction of HDR images. To address this problem, we propose a method to explicitly estimate the tone mapping function and its corresponding HDR image in one network. Firstly, based on the characteristics of the tone mapping function, we construct a model by a polynomial to describe the trend of the tone curve. To fit this curve, we use a learnable network to estimate the coefficients of the polynomial. This curve will be automatically adjusted according to the tone space of the Low Dynamic Range (LDR) image, and reconstruct the real HDR image. Besides, since all current datasets do not provide the corresponding relationship between the tone mapping function and the LDR image, we construct a new dataset with both synthetic and real images. Extensive experiments show that our method generalizes well under different tone-mapping functions and achieves SOTA performance. The code/dataset is available at https://github.com/jqtangust/EPCE-HDR.git.

Neural implicit surface learning has shown significant progress in multi-view 3D reconstruction, where an object is represented by multilayer perceptrons that provide continuous implicit surface representation and view-dependent radiance. However, current methods often fail to accurately reconstruct reflective surfaces, leading to severe ambiguity. To overcome this issue, we propose Ref-NeuS, which aims to reduce ambiguity by attenuating the effect of reflective surfaces. Specifically, we utilize an anomaly detector to estimate an explicit reflection score with the guidance of multiview context to localize reflective surfaces. Afterward, we design a reflection-aware photometric loss that adaptively reduces ambiguity by modeling rendered color as a Gaussian distribution, with the reflection score representing the variance. We show that together with a reflection direction-dependent radiance, our model achieves high-quality surface reconstruction on reflective surfaces and outperforms the state-of-the-arts by a large margin. Besides, our model is also comparable on general surfaces.

Conditional diffusion models have demonstrated impressive performance in image manipulation tasks. The general pipeline involves adding noise to the image and then denoising it. However, this method faces a trade-off problem: adding too much noise affects the fidelity of the image while adding too little affects its editability. This largely limits their practical applicability. In this paper, we propose a novel framework, Selective Diffusion Distillation (SDD), that ensures both the fidelity and editability of images. Instead of directly editing images with a diffusion model, we train a feedforward image manipulation network under the guidance of the diffusion model. Besides, we propose an effective indicator to select the semantic-related timestep to obtain the correct semantic guidance from the diffusion model. This approach successfully avoids the dilemma caused by the diffusion process. Our extensive experiments demonstrate the advantages of our framework. Code is released at https://github.com/AndysonYs/Selective-Diffusion-Distillation.

The fidelity of Generative Adversarial Networks (GAN) inversion is impeded by Out-Of-Domain (OOD) areas (e.g., background, accessories) in the image. Detecting the OOD areas beyond the generation ability of the pre-trained model and blending these regions with the input image can enhance fidelity. The "invertibility mask" figures out these OOD areas, and existing methods predict the mask with the reconstruction error. However, the estimated mask is usually inaccurate due to the influence of the reconstruction error in the In-Domain (ID) area. In this paper, we propose a novel framework that enhances the fidelity of human face in-version by designing a new module to decompose the input images to ID and OOD partitions with invertibility masks. Unlike previous works, our invertibility detector is simultaneously learned with a spatial alignment module. We iteratively align the generated features to the input geometry and reduce the reconstruction error in the ID regions. Thus, the OOD areas are more distinguishable and can be precisely predicted. Then, we improve the fidelity of our results by blending the OOD areas from the input image with the ID GAN inversion results. Our method produces photorealistic results for real-world human face image inversion and manipulation. Extensive experiments demonstrate our method’s superiority over existing methods in the quality of GAN inversion and attribute manipulation. Our code is available at: AbnerVictor/OOD-GAN-inversion

Remote photoplethysmography (rPPG) technology has become increasingly popular due to its non-invasive monitoring of various physiological indicators, making it widely applicable in multimedia interaction, healthcare, and emotion analysis. Existing rPPG methods utilize multiple datasets for training to enhance the generalizability of models. However, they often overlook the underlying conflict issues in the rPPG field, such as (1) label conflict resulting from different phase delays between physiological signal labels and face videos at the instance level, and (2) attribute conflict stemming from distribution shifts caused by head movements, illumination changes, skin types, etc. To address this, we introduce the DOmain-HArmonious framework (DOHA). Specifically, we first propose a harmonious phase strategy to eliminate uncertain phase delays and preserve the temporal variation of physiological signals. Next, we design a harmonious hyperplane optimization that reduces irrelevant attribute shifts and encourages the model's optimization towards a global solution that fits more valid scenarios. Our experiments demonstrate that DOHA significantly improves the performance of existing methods under multiple protocols.

Generalizable neural surface reconstruction techniques have attracted great attention in recent years. However, they encounter limitations of low confidence depth distribution and inaccurate surface reasoning due to the oversimplified volume rendering process employed. In this paper, we present Reconstruction TRansformer (ReTR), a novel framework that leverages the transformer architecture to redesign the rendering process, enabling complex render interaction modeling. It introduces a learnable $\textit{meta-ray token}$ and utilizes the cross-attention mechanism to simulate the interaction of rendering process with sampled points and render the observed color. Meanwhile, by operating within a high-dimensional feature space rather than the color space, ReTR mitigates sensitivity to projected colors in source views. Such improvements result in accurate surface assessment with high confidence. We demonstrate the effectiveness of our approach on various datasets, showcasing how our method outperforms the current state-of-the-art approaches in terms of reconstruction quality and generalization ability. $\textit{Our code is available at }$ https://github.com/YixunLiang/ReTR.

Abstract Neural radiance fields (NeRF) have demonstrated a promising research direction for novel view synthesis. However, the existing approaches either require per‐scene optimization that takes significant computation time or condition on local features which overlook the global context of images. To tackle this shortcoming, we propose the Conditionally Parameterized Neural Radiance Fields (CP‐NeRF), a plug‐in module that enables NeRF to leverage contextual information from different scales. Instead of optimizing the model parameters of NeRFs directly, we train a Feature Pyramid hyperNetwork (FPN) that extracts view‐dependent global and local information from images within or across scenes to produce the model parameters. Our model can be trained end‐to‐end with standard photometric loss from NeRF. Extensive experiments demonstrate that our method can significantly boost the performance of NeRF, achieving state‐of‐the‐art results in various benchmark datasets.

Deep neural networks suffer from significant performance deterioration when there exists distribution shift between deployment and training. Domain Generalization (DG) aims to safely transfer a model to unseen target domains by only relying on a set of source domains. Although various DG approaches have been proposed, a recent study named DomainBed (Gulrajani and Lopez-Paz, 2020), reveals that most of them do not beat simple empirical risk minimization (ERM). To this end, we propose a general framework that is orthogonal to existing DG algorithms and could improve their performance consistently. Unlike previous DG works that stake on a static source model to be hopefully a universal one, our proposed AdaODM adaptively modifies the source model at test time for different target domains. Specifically, we create multiple domain-specific classifiers upon a shared domain-generic feature extractor. The feature extractor and classifiers are trained in an adversarial way, where the feature extractor embeds the input samples into a domain-invariant space, and the multiple classifiers capture the distinct decision boundaries that each of them relates to a specific source domain. During testing, distribution differences between target and source domains could be effectively measured by leveraging prediction disagreement among source classifiers. By fine-tuning source models to minimize the disagreement at test time, target-domain features are well aligned to the invariant feature space. We verify AdaODM on two popular DG methods, namely ERM and CORAL, and four DG benchmarks, namely VLCS, PACS, OfficeHome, and TerraIncognita. The results show AdaODM stably improves the generalization capacity on unseen domains and achieves state-of-the-art performance.

In this work, we study the continual semantic segmentation problem, where the deep neural networks are required to incorporate new classes continually without catastrophic forgetting. We propose to use a structural re-parameterization mechanism, named representation compensation (RC) module, to decouple the representation learning of both old and new knowledge. The RC module consists of two dynamically evolved branches with one frozen and one trainable. Besides, we design a pooled cube knowledge distillation strategy on both spatial and channel dimensions to further enhance the plasticity and stability of the model. We conduct experiments on two challenging continual semantic segmentation scenarios, continual class segmentation and continual domain segmentation. Without any extra computational overhead and parameters during inference, our method outperforms state-of-the-art performance. The code is available at https://github.com/zhangchbin/RCIL.

Generative adversarial networks have achieved great success in unpaired image-to-image translation. Cycle consistency, a key component for this task, allows modeling the relationship between two distinct domains without paired data. In this paper, we propose an alternative framework, as an extension of latent space interpolation, to consider the intermediate region between two domains during translation. It is based on the assumption that in a flat and smooth latent space, there exist many paths that connect two sample points. Properly selecting paths makes it possible to change only certain image attributes, which is useful for generating intermediate images between the two domains. With this idea, our framework includes an encoder, an interpolator and a decoder. The encoder maps natural images to a convex and smooth latent space where interpolation is applicable. The interpolator controls the interpolation path so that desired intermediate samples can be obtained. Finally, the decoder inverts interpolated features back to pixel space. We also show that by choosing different reference images and interpolation paths, this framework can be applied to multi-domain and multi-modal translation. Extensive experiments manifest that our framework achieves superior results and is flexible for various tasks.

In this paper, we explore the mask representation in instance segmentation with Point-of-Interest (PoI) features. Differentiating multiple potential instances within a single PoI feature is challenging, because learning a high-dimensional mask feature for each instance using vanilla convolution demands a heavy computing burden. To address this challenge, we propose an instance-aware convolution. It decomposes this mask representation learning task into two tractable modules as instance-aware weights and instance-agnostic features. The former is to parametrize convolution for producing mask features corresponding to different instances, improving mask learning efficiency by avoiding employing several independent convolutions. Meanwhile, the latter serves as mask templates in a single point. Together, instance-aware mask features are computed by convolving the template with dynamic weights, used for the mask prediction. Along with instance-aware convolution, we propose PointINS, a simple and practical instance segmentation approach, building upon dense one-stage detectors. Through extensive experiments, we evaluated the effectiveness of our framework built upon RetinaNet and FCOS. PointINS in ResNet101 backbone achieves a 38.3 mask mean average precision (mAP) on COCO dataset, outperforming existing point-based methods by a large margin. It gives a comparable performance to the region-based Mask R-CNN K. He, G. Gkioxari, P. Dollár, and R. Girshick, "Mask R-CNN," in Proc. IEEE Int. Conf. Comput. Vis., 2017, pp. 2980-2988 with faster inference.

Text is a new way to guide human image manipulation. Albeit natural and flexible, text usually suffers from inaccuracy in spatial description, ambiguity in the description of appearance, and incompleteness. We in this paper address these issues. To overcome inaccuracy, we use structured information (e.g., poses) to help identify correct location to manipulate, by disentangling the control of appearance and spatial structure. Moreover, we learn the image-text shared space with derived disentanglement to improve accuracy and quality of manipulation, by separating relevant and irrelevant editing directions for the textual instructions in this space. Our model generates a series of manipulation results by moving source images in this space with different degrees of editing strength. Thus, to reduce the ambiguity in text, our model generates sequential output for manual selection. In addition, we propose an efficient pseudo-label loss to enhance editing performance when the text is incomplete. We evaluate our method on various datasets and show its precision and interactiveness to manipulate human images.

In this paper, we present a novel approach to synthesize realistic images based on their semantic layouts. It hypothesizes that for objects with similar appearance, they share similar representation. Our method establishes dependencies between regions according to their appearance correlation, yielding both spatially variant and associated representations. Conditioning on these features, we propose a dynamic weighted network constructed by spatially conditional computation (with both convolution and normalization). More than preserving semantic distinctions, the given dynamic network strengthens semantic relevance, benefiting global structure and detail synthesis. We demonstrate that our method gives the compelling generation performance qualitatively and quantitatively with extensive experiments on benchmarks.

Person re-identification (ReID) has gained an impressive progress in recent years. However, the occlusion is still a common and challenging problem for recent ReID methods. Several mainstream methods utilize extra cues (e.g., human pose information) to distinguish human parts from obstacles to alleviate the occlusion problem. Although achieving inspiring progress, these methods severely rely on the fine-grained extra cues, and are sensitive to the estimation error in the extra cues. In this paper, we show that existing methods may degrade if the extra information is sparse or noisy. Thus we propose a simple yet effective method that is robust to sparse and noisy pose information. This is achieved by discretizing pose information to the visibility label of body parts, so as to suppress the influence of occluded regions. We show in our experiments that leveraging pose information in this way is more effective and robust. Besides, our method can be embedded into most person ReID models easily. Extensive experiments validate the effectiveness of our model on common occluded person ReID datasets.

Real-world data often exhibit imbalanced distributions, where certain target values have significantly fewer observations. Existing techniques for dealing with imbalanced data focus on targets with categorical indices, i.e., different classes. However, many tasks involve continuous targets, where hard boundaries between classes do not exist. We define Deep Imbalanced Regression (DIR) as learning from such imbalanced data with continuous targets, dealing with potential missing data for certain target values, and generalizing to the entire target range. Motivated by the intrinsic difference between categorical and continuous label space, we propose distribution smoothing for both labels and features, which explicitly acknowledges the effects of nearby targets, and calibrates both label and learned feature distributions. We curate and benchmark large-scale DIR datasets from common real-world tasks in computer vision, natural language processing, and healthcare domains. Extensive experiments verify the superior performance of our strategies. Our work fills the gap in benchmarks and techniques for practical imbalanced regression problems. Code and data are available at: https://github.com/YyzHarry/imbalanced-regression.

Unpaired image-to-image translation (I2I) has achieved great success in various applications. However, its generalization capacity is still an open question. In this paper, we show that existing I2I models do not generalize well for samples outside the training domain. The cause is twofold. First, an I2I model may not work well when testing samples are beyond its valid input domain. Second, results could be unreliable if the expected output is far from what the model is trained. To deal with these issues, we propose the Domain Adaptive Image-To-Image translation (DAI2I) framework that adapts an I2I model for out-of-domain samples. Our framework introduces two sub-modules -- one maps testing samples to the valid input domain of the I2I model, and the other transforms the output of I2I model to expected results. Extensive experiments manifest that our framework improves the capacity of existing I2I models, allowing them to handle samples that are distinctively different from their primary targets.

Traditional convolution-based generative adversarial networks synthesize images based on hierarchical local operations, where long-range dependency relation is implicitly modeled with a Markov chain. It is still not sufficient for categories with complicated structures. In this paper, we characterize long-range dependence with attentive normalization (AN), which is an extension to traditional instance normalization. Specifically, the input feature map is softly divided into several regions based on its internal semantic similarity, which are respectively normalized. It enhances consistency between distant regions with semantic correspondence. Compared with self-attention GAN, our attentive normalization does not need to measure the correlation of all locations, and thus can be directly applied to large-size feature maps without much computational burden. Extensive experiments on class-conditional image generation and semantic inpainting verify the efficacy of our proposed module.

While video-based person re-identification (Re-ID) has drawn increasing attention and made great progress in recent years, it is still very challenging to effectively overcome the occlusion problem and the visual ambiguity problem for visually similar negative samples. On the other hand, we observe that different frames of a video can provide complementary information for each other, and the structural information of pedestrians can provide extra discriminative cues for appearance features. Thus, modeling the temporal relations of different frames and the spatial relations within a frame has the potential for solving the above problems. In this work, we propose a novel Spatial-Temporal Graph Convolutional Network (STGCN) to solve these problems. The STGCN includes two GCN branches, a spatial one and a temporal one. The spatial branch extracts structural information of a human body. The temporal branch mines discriminative cues from adjacent frames. By jointly optimizing these branches, our model extracts robust spatial-temporal information that is complementary with appearance information. As shown in the experiments, our model achieves state-of-the-art results on MARS and DukeMTMC-VideoReID datasets.

Deep neural network-based methods were proposed for face attribute manipulation. There still exist, however, two major issues, i.e., insufficient visual quality (or resolution) of the results and lack of user control. They limit the applicability of existing methods since users may have different editing preference on facial attributes. In this paper, we address these issues by proposing a semantic component model. The model decomposes a facial attribute into multiple semantic components, each corresponds to a specific face region. This not only allows for user control of edit strength on different parts based on their preference, but also makes it effective to remove unwanted edit effect. Further, each semantic component is composed of two fundamental elements, which determine the edit effect and region respectively. This property provides fine interactive control. As shown in experiments, our model not only produces high-quality results, but also allows effective user interaction.

Generative adversarial networks have achieved great success in unpaired image-to-image translation. Cycle consistency allows modeling the relationship between two distinct domains without paired data. In this paper, we propose an alternative framework, as an extension of latent space interpolation, to consider the intermediate region between two domains during translation. It is based on the fact that in a flat and smooth latent space, there exist many paths that connect two sample points. Properly selecting paths makes it possible to change only certain image attributes, which is useful for generating intermediate images between the two domains. We also show that this framework can be applied to multi-domain and multi-modal translation. Extensive experiments manifest its generality and applicability to various tasks.

Synthesizing novel views from a 2D image requires to infer 3D structure and project it back to 2D from a new viewpoint. In this paper, we propose an encoder-decoder based generative adversarial network VI-GAN to tackle this problem. Our method is to let the network, after seeing many images of objects belonging to the same category in different views, obtain essential knowledge of intrinsic properties of the objects. To this end, an encoder is designed to extract view-independent feature that characterizes intrinsic properties of the input image, which includes 3D structure, color, texture etc. We also make the decoder hallucinate the image of a novel view based on the extracted feature and an arbitrary user-specific camera pose. Extensive experiments demonstrate that our model can synthesize high-quality images in different views with continuous camera poses, and is general for various applications.

Digital face manipulation has become a popular and fascinating way to touch images with the prevalence of smart phones and social networks. With a wide variety of user preferences, facial expressions, and accessories, a general and flexible model is necessary to accommodate different types of facial editing. In this paper, we propose a model to achieve this goal based on an end-to-end convolutional neural network that supports fast inference, edit-effect control, and quick partial-model update. In addition, this model learns from unpaired image sets with different attributes. Experimental results show that our framework can handle a wide range of expressions, accessories, and makeup effects. It produces high-resolution and high-quality results in fast speed.

The challenge of person re-identification (re-id) is to match individual images of the same person captured by different non-overlapping camera views against significant and unknown cross-view feature distortion. While a large number of distance metric/subspace learning models have been developed for re-id, the cross-view transformations they learned are view-generic and thus potentially less effective in quantifying the feature distortion inherent to each camera view. Learning view-specific feature transformations for re-id (i.e., view-specific re-id), an under-studied approach, becomes an alternative resort for this problem. In this work, we formulate a novel view-specific person re-identification framework from the feature augmentation point of view, called Camera coR relation Aware Feature augmenTation (CRAFT). Specifically, CRAFT performs cross-view adaptation by automatically measuring camera correlation from cross-view visual data distribution and adaptively conducting feature augmentation to transform the original features into a new adaptive space. Through our augmentation framework, view-generic learning algorithms can be readily generalized to learn and optimize view-specific sub-models whilst simultaneously modelling view-generic discrimination information. Therefore, our framework not only inherits the strength of view-generic model learning but also provides an effective way to take into account view specific characteristics. Our CRAFT framework can be extended to jointly learn view-specific feature transformations for person re-id across a large network with more than two cameras, a largely under-investigated but realistic re-id setting. Additionally, we present a domain-generic deep person appearance representation which is designed particularly to be towards view invariant for facilitating cross-view adaptation by CRAFT. We conducted extensively comparative experiments to validate the superiority and advantages of our proposed framework over state-of-the-art competitors on contemporary challenging person re-id datasets.

Virtual face beautification (or markup) becomes common operations in camera or image processing Apps, which is actually deceiving. In this paper, we propose the task of restoring a portrait image from this process. As the first attempt along this line, we assume unknown global operations on human faces and aim to tackle the two issues of skin smoothing and skin color change. These two tasks, intriguingly, impose very different difficulties to estimate subtle details and major color variation. We propose a Component Regression Network (CRN) and address the limitation of using Euclidean loss in blind reversion. CRN maps the edited portrait images back to the original ones without knowing beautification operation details. Our experiments demonstrate effectiveness of the system for this novel task.

Person reidentification, which matches person images of the same identity across nonoverlapping camera views, becomes an important component for cross-camera-view activity analysis. Most (if not all) person reidentification algorithms are designed based on appearance features. However, appearance features are not stable across nonoverlapping camera views under dramatic lighting change, and those algorithms assume that two cross-view images of the same person can be well represented either by exploring robust and invariant features or by learning matching distance. Such an assumption ignores the nature that images are captured under different camera views with different camera characteristics and environments, and thus, mostly there exists large discrepancy between the extracted features under different views. To solve this problem, we formulate an asymmetric distance model for learning camera-specific projections to transform the unmatched features of each view into a common space where discriminative features across view space are extracted. A cross-view consistency regularization is further introduced to model the correlation between view-specific feature transformations of different camera views, which reflects their nature relations and plays a significant role in avoiding overfitting. A kernel cross-view discriminant component analysis is also presented. Extensive experiments have been conducted to show that asymmetric distance modeling is important for person reidentification, which matches the concerns on cross-disjoint-view matching, reporting superior performance compared with related distance learning methods on six publically available data sets.

Based on minimum reconstruction error criterion and the intrinsic sparse property of natural data, sparse representation (SR) has shown promising performance on various image recognition tasks. However, in the field of person re-identification (re-id), the state-of-the-art is still dominated by other methods such as metric learning or CNN. It is because samples in one view may not be representative enough to represent samples from another view. As such, the reconstruction error could be excessive, and different pedestrians are indistinguishable with the coefficient produced by sparse representation. In this paper, we proposed an asymmetric sparse representation to address this problem. Samples of different camera views (gallery and probe samples) are mapped to a common latent space and the sparse coefficient is generated in this space. In this way, the representation power is enhanced and the sparse coefficient becomes more reliable. The similarities of different samples are determined by the enhanced sparse coefficient, which allows more discriminative matching across different camera views. Extensive experiments on CAVIAR4REID, iLIDS-VID and PRID 2011 datasets have demonstrated the merits of our approach.

Feature representation and metric learning are two critical components in person re-identification models. In this paper, we focus on the feature representation and claim that hand-crafted histogram features can be complementary to Convolutional Neural Network (CNN) features. We propose a novel feature extraction model called Feature Fusion Net (FFN) for pedestrian image representation. In FFN, back propagation makes CNN features constrained by the handcrafted features. Utilizing color histogram features (RGB, HSV, YCbCr, Lab and YIQ) and texture features (multi-scale and multi-orientation Gabor features), we get a new deep feature representation that is more discriminative and compact. Experiments on three challenging datasets (VIPeR, CUHK01, PRID450s) validates the effectiveness of our proposal.