Weijie Wang

Bin Ren, Guofeng Mei, Danda Pani Paudel et al.

Contrastive learning (CL) for Vision Transformers (ViTs) in image domains has achieved performance comparable to CL for traditional convolutional backbones. However, in 3D point cloud pretraining with ViTs, masked autoencoder (MAE) modeling remains dominant. This raises the question: Can we take the best of both worlds? To answer this question, we first empirically validate that integrating MAE-based point cloud pre-training with the standard contrastive learning paradigm, even with meticulous design, can lead to a decrease in performance. To address this limitation, we reintroduce CL into the MAE-based point cloud pre-training paradigm by leveraging the inherent contrastive properties of MAE. Specifically, rather than relying on extensive data augmentation as commonly used in the image domain, we randomly mask the input tokens twice to generate contrastive input pairs. Subsequently, a weight-sharing encoder and two identically structured decoders are utilized to perform masked token reconstruction. Additionally, we propose that for an input token masked by both masks simultaneously, the reconstructed features should be as similar as possible. This naturally establishes an explicit contrastive constraint within the generative MAE-based pre-training paradigm, resulting in our proposed method, Point-CMAE. Consequently, Point-CMAE effectively enhances the representation quality and transfer performance compared to its MAE counterpart. Experimental evaluations across various downstream applications, including classification, part segmentation, and few-shot learning, demonstrate the efficacy of our framework in surpassing state-of-the-art techniques under standard ViTs and single-modal settings. The source code and trained models are available at: https://github.com/Amazingren/Point-CMAE.

Guofeng Mei, Hao Tang, Xiaoshui Huang et al.

Deep point cloud registration methods face challenges to partial overlaps and rely on labeled data. To address these issues, we propose UDPReg, an unsupervised deep probabilistic registration framework for point clouds with partial overlaps. Specifically, we first adopt a network to learn posterior probability distributions of Gaussian mixture models (GMMs) from point clouds. To handle partial point cloud registration, we apply the Sinkhorn algorithm to predict the distribution-level correspondences under the constraint of the mixing weights of GMMs. To enable unsupervised learning, we design three distribution consistency-based losses: self-consistency, cross-consistency, and local contrastive. The self-consistency loss is formulated by encouraging GMMs in Euclidean and feature spaces to share identical posterior distributions. The cross-consistency loss derives from the fact that the points of two partially overlapping point clouds belonging to the same clusters share the cluster centroids. The cross-consistency loss allows the network to flexibly learn a transformation-invariant posterior distribution of two aligned point clouds. The local contrastive loss facilitates the network to extract discriminative local features. Our UDPReg achieves competitive performance on the 3DMatch/3DLoMatch and ModelNet/ModelLoNet benchmarks.

Songlong Xing, Weijie Wang, Zhengyu Zhao et al.

Despite their impressive zero-shot abilities, vision-language models such as CLIP have been shown to be susceptible to adversarial attacks. To enhance its adversarial robustness, recent studies finetune the pretrained vision encoder of CLIP with adversarial examples on a proxy dataset such as ImageNet by aligning adversarial images with correct class labels. However, these methods overlook the important roles of training data distributions and learning objectives, resulting in reduced zero-shot capabilities and limited transferability of robustness across domains and datasets. In this work, we propose a simple yet effective paradigm AdvFLYP, which follows the training recipe of CLIP's pretraining process when performing adversarial finetuning to the model. Specifically, AdvFLYP finetunes CLIP with adversarial images created based on image-text pairs collected from the web, and match them with their corresponding texts via a contrastive loss. To alleviate distortion of adversarial image embeddings of noisy web images, we further propose to regularise AdvFLYP by penalising deviation of adversarial image features. We show that logit- and feature-level regularisation terms benefit robustness and clean accuracy, respectively. Extensive experiments on 14 downstream datasets spanning various domains show the superiority of our paradigm over mainstream practices. Our code and model weights are released at https://github.com/Sxing2/AdvFLYP.

Mingrui Ma, Weijie Wang, Jie Ning et al.

The challenge of Multimodal Deformable Image Registration (MDIR) lies in the conversion and alignment of features between images of different modalities. Generative models (GMs) cannot retain the necessary information enough from the source modality to the target one, while non-GMs struggle to align features across these two modalities. In this paper, we propose a novel coarse-to-fine MDIR framework,LLM-Morph, which is applicable to various pre-trained Large Language Models (LLMs) to solve these concerns by aligning the deep features from different modal medical images. Specifically, we first utilize a CNN encoder to extract deep visual features from cross-modal image pairs, then we use the first adapter to adjust these tokens, and use LoRA in pre-trained LLMs to fine-tune their weights, both aimed at eliminating the domain gap between the pre-trained LLMs and the MDIR task. Third, for the alignment of tokens, we utilize other four adapters to transform the LLM-encoded tokens into multi-scale visual features, generating multi-scale deformation fields and facilitating the coarse-to-fine MDIR task. Extensive experiments in MR-CT Abdomen and SR-Reg Brain datasets demonstrate the effectiveness of our framework and the potential of pre-trained LLMs for MDIR task. Our code is availabel at: https://github.com/ninjannn/LLM-Morph.

Xiaorong Zhu, Qiang Li, Zibo Xu et al.

Medical visual question answering requires models to ground their responses in image evidence, because visually unsupported answers can mislead downstream interpretation. However, many medical VQA questions are generic, template-like, or highly similar in form, which can encourage models to learn question-answer shortcuts instead of image-dependent reasoning and thereby increase the risk of hallucinated responses. We propose Ask4VG, a label-free pilot framework for risk-aware question selection. Ask4VG estimates question-induced hallucination risk through counterfactual visual probing: the same question is asked under the original image, a perturbed image, a blank image, and a mismatched image, and the resulting answer relations are converted into weak supervision for a counterfactual risk estimator. The learned estimator then reranks candidate question rewrites to favor intent-preserving questions that are less invariant to missing or mismatched visual evidence before final answer generation. On VQA-RAD with Qwen2-VL-2B-Instruct, prompt-only rewriting increases counterfactual risk, whereas predicted-risk reranking reduces held-out risk from 0.658 to 0.623 and improves exact accuracy from 0.337 to 0.356. A 300-sample PMC-VQA external check shows the same direction of risk reduction with a small accuracy gain. These results suggest that question selection is a promising complement to response-level hallucination mitigation for reliable medical VQA.

Yuling Jiao, Qinian Jin, Xiliang Lu et al.

In this paper we propose an iterative method using alternating direction method of multipliers (ADMM) strategy to solve linear inverse problems in Hilbert spaces with general convex penalty term. When the data is given exactly, we give a convergence analysis of our ADMM algorithm without assuming the existence of Lagrange multiplier. In case the data contains noise, we show that our method is a regularization method as long as it is terminated by a suitable stopping rule. Various numerical simulations are performed to test the efficiency of the method.

Weijie Wang, Qihang Cao, Sensen Gao et al.

Reconstructing 3D representations from 2D inputs is a fundamental task in computer vision and graphics, serving as a cornerstone for understanding and interacting with the physical world. While traditional methods achieve high fidelity, they are limited by slow per-scene optimization or category-specific training, which hinders their practical deployment and scalability. Hence, generalizable feed-forward 3D reconstruction has witnessed rapid development in recent years. By learning a model that maps images directly to 3D representations in a single forward pass, these methods enable efficient reconstruction and robust cross-scene generalization. Our survey is motivated by a critical observation: despite the diverse geometric output representations, ranging from implicit fields to explicit primitives, existing feed-forward approaches share similar high-level architectural patterns, such as image feature extraction backbones, multi-view information fusion mechanisms, and geometry-aware design principles. Consequently, we abstract away from these representation differences and instead focus on model design, proposing a novel taxonomy centered on model design strategies that are agnostic to the output format. Our proposed taxonomy organizes the research directions into five key problems that drive recent research development: feature enhancement, geometry awareness, model efficiency, augmentation strategies and temporal-aware models. To support this taxonomy with empirical grounding and standardized evaluation, we further comprehensively review related benchmarks and datasets, and extensively discuss and categorize real-world applications based on feed-forward 3D models. Finally, we outline future directions to address open challenges such as scalability, evaluation standards, and world modeling.

Haoyu Zhao, Akide Liu, Zeyu Zhang et al.

Embodied question answering (EQA) in 3D environments often requires collecting context that is distributed across multiple viewpoints and partially occluded. However, most recent vision--language models (VLMs) are constrained to a fixed and finite set of input views, which limits their ability to acquire question-relevant context at inference time and hinders complex spatial reasoning. We propose Chain-of-View (CoV) prompting, a training-free, test-time reasoning framework that transforms a VLM into an active viewpoint reasoner through a coarse-to-fine exploration process. CoV first employs a View Selection agent to filter redundant frames and identify question-aligned anchor views. It then performs fine-grained view adjustment by interleaving iterative reasoning with discrete camera actions, obtaining new observations from the underlying 3D scene representation until sufficient context is gathered or a step budget is reached. We evaluate CoV on OpenEQA across four mainstream VLMs and obtain an average +11.56% improvement in LLM-Match, with a maximum gain of +13.62% on Qwen3-VL-Flash. CoV further exhibits test-time scaling: increasing the minimum action budget yields an additional +2.51% average improvement, peaking at +3.73% on Gemini-2.5-Flash. On ScanQA and SQA3D, CoV delivers strong performance (e.g., 116 CIDEr / 31.9 EM@1 on ScanQA and 51.1 EM@1 on SQA3D). Overall, these results suggest that question-aligned view selection coupled with open-view search is an effective, model-agnostic strategy for improving spatial reasoning in 3D EQA without additional training. Code is available on https://github.com/ziplab/CoV .

Weijie Wang, Bo Li, Nicu Sebe et al.

Due to the subjective crowdsourcing annotations and the inherent inter-class similarity of facial expressions, the real-world Facial Expression Recognition (FER) datasets usually exhibit ambiguous annotation. To simplify the learning paradigm, most previous methods convert ambiguous annotation results into precise one-hot annotations and train FER models in an end-to-end supervised manner. In this paper, we rethink the existing training paradigm and propose that it is better to use weakly supervised strategies to train FER models with original ambiguous annotation.

Akide Liu, Jinbo Xing, Chaojie Mao et al.

Minute-scale cinematic video generation is a central challenge for generative video models. Existing paradigms address only fragments of this challenge: single-shot extrapolation preserves an anchor but lacks cinematic structure, while multi-shot storytelling imposes structure yet remains free to invent its visual states rather than continue an observed one. We define Multi-Shot Video Extrapolation (MSVE), a task that extends an observed frame or clip into a sequence of cinematically structured shots while preserving anchor state and advancing narrative intent. This setting operates under the finite per-call generation budget of short-video models. We identify three coupled bottlenecks: (1) global planners over-specify unsupported details from full screenplays; (2) shot-level prompts dilute task-relevant state when carrying the complete story; and (3) temporal chaining turns generated frames into a lossy memory in which identity, scene, object, and action state decay. MSVE reveals that long-video failure is not merely a limitation of context length, but a failure of context allocation. We propose Recursive Context Allocation (ReCA), an inference-time framework that allocates context hierarchically across planning and generation. ReCA recursively decomposes MSVE into context-bounded subproblems, invokes frozen generators at leaf nodes, and propagates structured state updates across time. To evaluate this setting, we further propose MSVE-Bench and NB-Q, a source-grounded protocol with prompts purpose-built for 3 to 5 minute long-video generation, a regime not addressed by existing short-clip benchmarks. Compared to previous methods, ReCA improves average normalized score by 8 to 16 percent over the strongest competing controller and improves multi-shot consistency metrics by 28 to 43 percent. View the project page at https://reca.vmv.re.

Weijie Wang, Zimu Li, Jinchuan Shi et al.

Sparse-view 3D reconstruction is increasingly addressed with feed-forward splatting networks that predict explicit primitives directly from images. Yet most existing methods remain centered on Gaussian primitives and expose surfaces only indirectly: extracting a usable mesh for downstream simulation, physics reasoning, or embodied interaction still requires expensive post-hoc steps that break the feed-forward promise. This limitation is especially pronounced in pose-free settings, where scene structure and camera parameters must be estimated jointly from sparse observations. We present TriSplat, a feed-forward reconstruction network that represents scenes with oriented triangle primitives and directly exports simulation-ready mesh scenes from a single forward pass. Given input images, the network predicts local 3D point maps, triangle attributes, camera poses, and optional intrinsics. Rather than regressing triangle orientation as an unconstrained latent variable, our approach constructs geometry normals from the predicted point maps, refines them with an image-conditioned normal head, and converts them into stable local frames for triangle parameterization. A mono-normal bootstrap schedule further stabilizes early training, while opacity and blur scheduling progressively sharpens the learned surface representation for direct mesh extraction. Experiments on RealEstate10K and DL3DV show that this representation produces more geometry-faithful reconstructions than Gaussian feed-forward baselines while maintaining competitive novel-view rendering quality. Because the rendering primitives are themselves surface triangles, the output can be directly ingested by physics engines, collision detectors, and standard rendering pipelines without any conversion, making it a practical simulation-ready solution for feed-forward 3D scene reconstruction.

Andrea Rigo, Luca Stornaiuolo, Weijie Wang et al.

We propose a diffusion-based approach for Text-to-Image (T2I) generation with consistent and interactive 3D layout control and editing. While prior methods improve spatial adherence using 2D cues or iterative copy-warp-paste strategies, they often distort object geometry and fail to preserve consistency across edits. To address these limitations, we introduce a framework for Positioning Objects Consistently and Interactively (POCI-Diff), a novel formulation for jointly enforcing 3D geometric constraints and instance-level semantic binding within a unified diffusion process. Our method enables explicit per-object semantic control by binding individual text descriptions to specific 3D bounding boxes through Blended Latent Diffusion, allowing one-shot synthesis of complex multi-object scenes. We further propose a warping-free generative editing pipeline that supports object insertion, removal, and transformation via regeneration rather than pixel deformation. To preserve object identity and consistency across edits, we condition the diffusion process on reference images using IP-Adapter, enabling coherent object appearance throughout interactive 3D editing while maintaining global scene coherence. Experimental results demonstrate that POCI-Diff produces high-quality images consistent with the specified 3D layouts and edits, outperforming state-of-the-art methods in both visual fidelity and layout adherence while eliminating warping-induced geometric artifacts.

Shuang Chen, Yue Guo, Zhaochen Su et al.

Inspired by the remarkable reasoning capabilities of Deepseek-R1 in complex textual tasks, many works attempt to incentivize similar capabilities in Multimodal Large Language Models (MLLMs) by directly applying reinforcement learning (RL). However, they still struggle to activate complex reasoning. In this paper, rather than examining multimodal RL in isolation, we delve into current training pipelines and identify three crucial phenomena: 1) Effective cold start initialization is critical for enhancing MLLM reasoning. Intriguingly, we find that initializing with carefully selected text data alone can lead to performance surpassing many recent multimodal reasoning models, even before multimodal RL. 2) Standard GRPO applied to multimodal RL suffers from gradient stagnation, which degrades training stability and performance. 3) Subsequent text-only RL training, following the multimodal RL phase, further enhances multimodal reasoning. This staged training approach effectively balances perceptual grounding and cognitive reasoning development. By incorporating the above insights and addressing multimodal RL issues, we introduce ReVisual-R1, achieving a new state-of-the-art among open-source 7B MLLMs on challenging benchmarks including MathVerse, MathVision, WeMath, LogicVista, DynaMath, and challenging AIME2024 and AIME2025.

Baojun Che, Yifan Chen, Daniel Zhengyu Huang et al.

Black-box variational inference (BBVI) with Gaussian mixture families offers a flexible approach for approximating complex posterior distributions without requiring gradients of the target density. However, standard numerical optimization methods often suffer from instability and inefficiency. We develop a stable and efficient framework that combines three key components: (1) affine-invariant preconditioning via natural gradient formulations, (2) an exponential integrator that unconditionally preserves the positive definiteness of covariance matrices, and (3) adaptive time stepping to ensure stability and to accommodate distinct warm-up and convergence phases. The proposed approach has natural connections to manifold optimization and mirror descent. For Gaussian posteriors, we prove exponential convergence in the noise-free setting and almost-sure convergence under Monte Carlo estimation, rigorously justifying the necessity of adaptive time stepping. Numerical experiments on multimodal distributions, Neal's multiscale funnel, and a PDE-based Bayesian inverse problem for Darcy flow demonstrate the effectiveness of the proposed method.

ZhiYuan Feng, Yu Deng, Ruichuan An et al.

In real home deployments, household agents must often operate from a complete household scene and a situated household request, rather than from a clean task specification. Such requests require agents to identify task-relevant entities, recover intended task conditions, and resolve ordering constraints from the surrounding scene context. We formalize this capability as full-scene household reasoning: given a complete household scene and a situated household request, an agent must infer executable task structure before producing a grounded skill-level action sequence. This setting is challenging because complete household scenes contain substantial task-irrelevant information, making direct complete-scene prompting inefficient and error-prone. In practical deployment, this challenge is further amplified by privacy and local compute constraints, which favor compact open-weight models with limited long-context reasoning ability. We propose TaskGround, a training-free and model-agnostic Ground-Infer-Execute framework that grounds complete scenes into compact task-relevant scene slices, infers executable task structure, and compiles it into grounded skill-level action sequences. To evaluate this setting, we introduce FullHome, a human-validated evaluation suite of 400 household tasks spanning diverse home-scale environments and both goal-oriented and process-constrained requirements. On FullHome, TaskGround improves task success rates by large margins across both proprietary and open-weight models. Notably, it makes Qwen3.5-9B competitive with GPT-5 under direct complete-scene prompting while reducing total input-token cost by up to 18x. Our results identify executable task-structure inference as a central bottleneck in full-scene household reasoning and show that structured grounding can make compact local models substantially more effective for practical household deployment.

Weijie Wang, Jichao Zhang, Chang Liu et al.

Recently, diffusion models have made significant strides in synthesizing realistic 2D human images based on provided text prompts. Building upon this, researchers have extended 2D text-to-image diffusion models into the 3D domain for generating human textures (UV Maps). However, some important problems about UV Map Generative models are still not solved, i.e., how to generate personalized texture maps for any given face image, and how to define and evaluate the quality of these generated texture maps. To solve the above problems, we introduce a novel method, UVMap-ID, which is a controllable and personalized UV Map generative model. Unlike traditional large-scale training methods in 2D, we propose to fine-tune a pre-trained text-to-image diffusion model which is integrated with a face fusion module for achieving ID-driven customized generation. To support the finetuning strategy, we introduce a small-scale attribute-balanced training dataset, including high-quality textures with labeled text and Face ID. Additionally, we introduce some metrics to evaluate the multiple aspects of the textures. Finally, both quantitative and qualitative analyses demonstrate the effectiveness of our method in controllable and personalized UV Map generation. Code is publicly available via https://github.com/twowwj/UVMap-ID.

Weijie Wang, Songlong Xing, Zhengyu Zhao et al.

Poisoning input views of 3D reconstruction systems has been recently studied. However, we identify that existing studies simply backpropagate adversarial gradients through the 3D reconstruction pipeline as a whole, without uncovering the new vulnerability rooted in specific modules of the 3D reconstruction pipeline. In this paper, we argue that the structure-from-motion (SfM) initialization, as the geometric core of many widely used reconstruction systems, can be targeted to achieve transferable poisoning effects across diverse 3D reconstruction systems. To this end, we propose PoInit-of-View, which optimizes adversarial perturbations to intentionally introduce cross-view gradient inconsistencies at projections of corresponding 3D points. These inconsistencies disrupt keypoint detection and feature matching, thereby corrupting pose estimation and triangulation within SfM, eventually resulting in low-quality rendered views. We also provide a theoretical analysis that connects cross-view inconsistency to correspondence collapse. Experimental results demonstrate the effectiveness of our PoInit-of-View on diverse 3D reconstruction systems and datasets, surpassing the single-view baseline by 25.1% in PSNR and 16.5% in SSIM in black-box transfer settings, such as 3DGS to NeRF.

Xiaoxuan He, Siming Fu, Zeyue Xue et al.

Group Relative Policy Optimization has emerged as essential for aligning video diffusion models with human preferences, but faces a critical computational bottleneck: training a 14B parametered model typically demands hundreds of GPU days per experiment. Existing efficiency methods reduce costs through sliding window subsampling training timesteps, but fundamentally compromise optimization, exhibiting severe instability and failing to reach full trajectory performance. We present Flash-GRPO, a single-step training framework that outperforms full trajectory training in alignment quality under low computational budgets while substantially improving training efficiency. Flash-GRPO addresses two critical challenges: iso-temporal grouping eliminates timestep-confounded variance by enforcing prompt-wise temporal consistency, decoupling policy performance from timestep difficulty; temporal gradient rectification neutralizes the time-dependent scaling factor that causes vastly inconsistent gradient magnitudes across timesteps. Experiments on 1.3B to 14B parameter models validate Flash-GRPO's effectiveness, demonstrating substantial training acceleration with consistent stability and state-of-the-art alignment quality.

Yunzhi Yan, Haotong Lin, Chenxu Zhou et al.

This paper aims to tackle the problem of modeling dynamic urban streets for autonomous driving scenes. Recent methods extend NeRF by incorporating tracked vehicle poses to animate vehicles, enabling photo-realistic view synthesis of dynamic urban street scenes. However, significant limitations are their slow training and rendering speed. We introduce Street Gaussians, a new explicit scene representation that tackles these limitations. Specifically, the dynamic urban scene is represented as a set of point clouds equipped with semantic logits and 3D Gaussians, each associated with either a foreground vehicle or the background. To model the dynamics of foreground object vehicles, each object point cloud is optimized with optimizable tracked poses, along with a 4D spherical harmonics model for the dynamic appearance. The explicit representation allows easy composition of object vehicles and background, which in turn allows for scene editing operations and rendering at 135 FPS (1066 $\times$ 1600 resolution) within half an hour of training. The proposed method is evaluated on multiple challenging benchmarks, including KITTI and Waymo Open datasets. Experiments show that the proposed method consistently outperforms state-of-the-art methods across all datasets. The code will be released to ensure reproducibility.

Zhiyuan Feng, Zhaolu Kang, Qijie Wang et al.

Vision-language models (VLMs) are essential to Embodied AI, enabling robots to perceive, reason, and act in complex environments. They also serve as the foundation for the recent Vision-Language-Action (VLA) models. Yet most evaluations of VLMs focus on single-view settings, leaving their ability to integrate multi-view information underexplored. At the same time, multi-camera setups are increasingly standard in robotic platforms, as they provide complementary perspectives to mitigate occlusion and depth ambiguity. Whether VLMs can effectively leverage such multi-view inputs for robotic reasoning therefore remains an open question. To bridge this gap, we introduce MV-RoboBench, a benchmark specifically designed to evaluate the multi-view spatial reasoning capabilities of VLMs in robotic manipulation. MV-RoboBench consists of 1.7k manually curated QA items across eight subtasks, divided into two primary categories: spatial understanding and robotic execution. We evaluate a diverse set of existing VLMs, including both open-source and closed-source models, along with enhanced versions incorporating CoT-inspired techniques. The results show that state-of-the-art models remain far below human performance, underscoring the substantial challenges VLMs face in multi-view robotic perception. Additionally, our analysis uncovers two key findings: (i) spatial intelligence and robotic task execution are positively correlated in multi-view robotic scenarios; and (ii) strong performance on existing general-purpose single-view spatial understanding benchmarks does not reliably translate to success in the robotic spatial tasks assessed by our benchmark. We release MV-RoboBench as an open resource to foster progress in spatially grounded VLMs and VLAs, providing not only data but also a standardized evaluation protocol for multi-view embodied reasoning.

Kehua Feng, Xinyi Shen, Weijie Wang et al.

Large language models (LLMs) are playing an increasingly important role in scientific research, yet there remains a lack of comprehensive benchmarks to evaluate the breadth and depth of scientific knowledge embedded in these models. To address this gap, we introduce SciKnowEval, a large-scale dataset designed to systematically assess LLMs across five progressive levels of scientific understanding: memory, comprehension, reasoning, discernment, and application. SciKnowEval comprises 28K multi-level questions and solutions spanning biology, chemistry, physics, and materials science. Using this benchmark, we evaluate 20 leading open-source and proprietary LLMs. The results show that while proprietary models often achieve state-of-the-art performance, substantial challenges remain -- particularly in scientific reasoning and real-world application. We envision SciKnowEval as a standard benchmark for evaluating scientific capabilities in LLMs and as a catalyst for advancing more capable and reliable scientific language models.

Weijie Wang, Zhengyu Zhao, Nicu Sebe et al.

Malicious use of deepfakes leads to serious public concerns and reduces people's trust in digital media. Although effective deepfake detectors have been proposed, they are substantially vulnerable to adversarial attacks. To evaluate the detector's robustness, recent studies have explored various attacks. However, all existing attacks are limited to 2D image perturbations, which are hard to translate into real-world facial changes. In this paper, we propose adversarial head turn (AdvHeat), the first attempt at 3D adversarial face views against deepfake detectors, based on face view synthesis from a single-view fake image. Extensive experiments validate the vulnerability of various detectors to AdvHeat in realistic, black-box scenarios. For example, AdvHeat based on a simple random search yields a high attack success rate of 96.8% with 360 searching steps. When additional query access is allowed, we can further reduce the step budget to 50. Additional analyses demonstrate that AdvHeat is better than conventional attacks on both the cross-detector transferability and robustness to defenses. The adversarial images generated by AdvHeat are also shown to have natural looks. Our code, including that for generating a multi-view dataset consisting of 360 synthetic views for each of 1000 IDs from FaceForensics++, is available at https://github.com/twowwj/AdvHeaT.

Mingrui Ma, Tao Wang, Lei Song et al.

The Swin transformer has recently attracted attention in medical image analysis due to its computational efficiency and long-range modeling capability. Owing to these properties, the Swin Transformer is suitable for establishing more distant relationships between corresponding voxels in different positions in complex abdominal image registration tasks. However, the registration models based on transformers combine multiple voxels into a single semantic token. This merging process limits the transformers to model and generate coarse-grained spatial information. To address this issue, we propose Recovery Feature Resolution Network (RFRNet), which allows the transformer to contribute fine-grained spatial information and rich semantic correspondences to higher resolution levels. Furthermore, shifted window partitioning operations are inflexible, indicating that they cannot perceive the semantic information over uncertain distances and automatically bridge the global connections between windows. Therefore, we present a Weighted Window Attention (WWA) to build global interactions between windows automatically. It is implemented after the regular and cyclic shift window partitioning operations within the Swin transformer block. The proposed unsupervised deformable image registration model, named RFR-WWANet, detects the long-range correlations, and facilitates meaningful semantic relevance of anatomical structures. Qualitative and quantitative results show that RFR-WWANet achieves significant improvements over the current state-of-the-art methods. Ablation experiments demonstrate the effectiveness of the RFRNet and WWA designs. Our code is available at \url{https://github.com/MingR-Ma/RFR-WWANet}.

Weizhi Nie, Zhen Qu, Weijie Wang et al.

Timely and interpretable early warning of sepsis remains a major clinical challenge due to the complex temporal dynamics of physiological deterioration. Traditional data-driven models often provide accurate yet opaque predictions, limiting physicians' confidence and clinical applicability. To address this limitation, we propose a Large Language Model (LLM)-guided temporal simulation framework that explicitly models physiological trajectories prior to disease onset for clinically interpretable prediction. The framework consists of a spatiotemporal feature extraction module that captures dynamic dependencies among multivariate vital signs, a Medical Prompt-as-Prefix module that embeds clinical reasoning cues into LLMs, and an agent-based post-processing component that constrains predictions within physiologically plausible ranges. By first simulating the evolution of key physiological indicators and then classifying sepsis onset, our model offers transparent prediction mechanisms that align with clinical judgment. Evaluated on the MIMIC-IV and eICU databases, the proposed method achieves superior AUC scores (0.861-0.903) across 24-4-hour pre-onset prediction tasks, outperforming conventional deep learning and rule-based approaches. More importantly, it provides interpretable trajectories and risk trends that can assist clinicians in early intervention and personalized decision-making in intensive care environments.

Junkang Zhou, Yefei He, Feng Chen et al.

Large-scale autoregressive models have demonstrated remarkable capabilities in image generation. However, their sequential raster-scan decoding relies on strictly next-token prediction, making inference prohibitively expensive. Existing acceleration methods typically either introduce entirely new generation paradigms that necessitate costly pre-training from scratch, or enable parallel generation at the expense of a training-inference gap or altered prediction objectives. In this paper, we introduce FlashAR, a lightweight post-training adaptation framework that efficiently adapts a pre-trained raster-scan autoregressive model into a highly parallel generator based on two-way next-token prediction. Our key insight is that effective adaptation should minimize modifications to the pre-trained model's original training objective to preserve its learned prior. Accordingly, we retain the original AR head as a horizontal head for row-wise prediction and introduce a complementary, lightweight vertical head for column-wise prediction. To facilitate efficient adaptation, we branch the vertical head from an intermediate layer rather than the final layer, bypassing the inherent horizontal head bias. Moreover, since horizontal and vertical predictions capture complementary dependencies whose relative importance varies across target positions, we employ a learnable fusion gate to dynamically combine the two predictions at each position. To further reduce adaptation cost, we propose a two-stage adaptation pipeline: the vertical head is first initialized through adaptation from the pre-trained autoregressive model before jointly fine-tuned with backbone to adapt to the new decoding paradigm. Extensive experiments on LlamaGen and Emu3.5 show that FlashAR achieves up to a 22.9x speedup for 512x512 image generation through a lightweight post-training with merely 0.05% of the original training data.

Jin Li, Tao Chen, Shuai Jiang et al.

Recent advances in diffusion models have significantly elevated the visual fidelity of Virtual Try-On (VTON) systems, yet reliable evaluation remains a persistent bottleneck. Traditional metrics struggle to quantify fine-grained texture details and semantic consistency, while existing datasets fail to meet commercial standards in scale and diversity. We present OpenVTON-Bench, a large-scale benchmark comprising approximately 100K high-resolution image pairs (up to $1536 \times 1536$). The dataset is constructed using DINOv3-based hierarchical clustering for semantically balanced sampling and Gemini-powered dense captioning, ensuring a uniform distribution across 20 fine-grained garment categories. To support reliable evaluation, we propose a multi-modal protocol that measures VTON quality along five interpretable dimensions: background consistency, identity fidelity, texture fidelity, shape plausibility, and overall realism. The protocol integrates VLM-based semantic reasoning with a novel Multi-Scale Representation Metric based on SAM3 segmentation and morphological erosion, enabling the separation of boundary alignment errors from internal texture artifacts. Experimental results show strong agreement with human judgments (Kendall's $τ$ of 0.833 vs. 0.611 for SSIM), establishing a robust benchmark for VTON evaluation.

Xiaolong Li, Youping Gu, Xi Lin et al.

Attention mechanisms are the core of foundation models, but their quadratic complexity remains a critical bottleneck for scaling. This challenge has driven the development of efficient attention mechanisms, with sparsity emerging as the dominant paradigm. Current methods typically retain or discard entire key-value blocks with binary masks, resulting in substantial information loss under high sparsity. To mitigate this gap, we present Pyramid Sparse Attention (PSA), a versatile module applicable to both video understanding and generation tasks. Instead of binary masking, PSA introduces multi-level pooled KV representations, enabling finer mask granularity. Specifically, each query block dynamically allocates lower pooling levels to critical KV blocks and higher levels to less important ones, creating an informative interpolation between full retention and complete pruning. This design, analogous to fixed-point quantization and classical feature pyramid networks in computer vision, effectively mitigates information loss while preserving computational efficiency under a low compute budget. It works with a native, hardware-friendly kernel that leverages decoupled block-tile design to ensure efficient execution. Across video understanding and generation benchmarks, PSA preserves contextual information and visual fidelity, consistently outperforming or achieving comparable performance over existing sparse attention baselines with superior efficiency-quality trade-offs. Our code and model weights are publicly available at: http://ziplab.co/PSA

Shuhong Liu, Chenyu Bao, Ziteng Cui et al.

This paper presents a comprehensive review of the NTIRE 2026 3D Restoration and Reconstruction (3DRR) Challenge, detailing the proposed methods and results. The challenge seeks to identify robust reconstruction pipelines that are robust under real-world adverse conditions, specifically extreme low-light and smoke-degraded environments, as captured by our RealX3D benchmark. A total of 279 participants registered for the competition, of whom 33 teams submitted valid results. We thoroughly evaluate the submitted approaches against state-of-the-art baselines, revealing significant progress in 3D reconstruction under adverse conditions. Our analysis highlights shared design principles among top-performing methods and provides insights into effective strategies for handling 3D scene degradation.

Weijie Wang, Xiaoxuan He, Youping Gu et al.

Recent video foundation models demonstrate impressive visual synthesis but frequently suffer from geometric inconsistencies. While existing methods attempt to inject 3D priors via architectural modifications, they often incur high computational costs and limit scalability. We propose World-R1, a framework that aligns video generation with 3D constraints through reinforcement learning. To facilitate this alignment, we introduce a specialized pure text dataset tailored for world simulation. Utilizing Flow-GRPO, we optimize the model using feedback from pre-trained 3D foundation models and vision-language models to enforce structural coherence without altering the underlying architecture. We further employ a periodic decoupled training strategy to balance rigid geometric consistency with dynamic scene fluidity. Extensive evaluations reveal that our approach significantly enhances 3D consistency while preserving the original visual quality of the foundation model, effectively bridging the gap between video generation and scalable world simulation.

Chaojun Ni, Xiaofeng Wang, Zheng Zhu et al.

Interactive 3D generation is gaining momentum and capturing extensive attention for its potential to create immersive virtual experiences. However, a critical challenge in current 3D generation technologies lies in achieving real-time interactivity. To address this issue, we introduce WonderTurbo, the first real-time interactive 3D scene generation framework capable of generating novel perspectives of 3D scenes within 0.72 seconds. Specifically, WonderTurbo accelerates both geometric and appearance modeling in 3D scene generation. In terms of geometry, we propose StepSplat, an innovative method that constructs efficient 3D geometric representations through dynamic updates, each taking only 0.26 seconds. Additionally, we design QuickDepth, a lightweight depth completion module that provides consistent depth input for StepSplat, further enhancing geometric accuracy. For appearance modeling, we develop FastPaint, a 2-steps diffusion model tailored for instant inpainting, which focuses on maintaining spatial appearance consistency. Experimental results demonstrate that WonderTurbo achieves a remarkable 15X speedup compared to baseline methods, while preserving excellent spatial consistency and delivering high-quality output.

Yuxing Chen, Weijie Wang, Sylvain Lobry et al.

Large language models (LLMs) are being used in data science code generation tasks, but they often struggle with complex sequential tasks, leading to logical errors. Their application to geospatial data processing is particularly challenging due to difficulties in incorporating complex data structures and spatial constraints, effectively utilizing diverse function calls, and the tendency to hallucinate less-used geospatial libraries. To tackle these problems, we introduce GeoAgent, a new interactive framework designed to help LLMs handle geospatial data processing more effectively. GeoAgent pioneers the integration of a code interpreter, static analysis, and Retrieval-Augmented Generation (RAG) techniques within a Monte Carlo Tree Search (MCTS) algorithm, offering a novel approach to geospatial data processing. In addition, we contribute a new benchmark specifically designed to evaluate the LLM-based approach in geospatial tasks. This benchmark leverages a variety of Python libraries and includes both single-turn and multi-turn tasks such as data acquisition, data analysis, and visualization. By offering a comprehensive evaluation among diverse geospatial contexts, this benchmark sets a new standard for developing LLM-based approaches in geospatial data analysis tasks. Our findings suggest that relying solely on knowledge of LLM is insufficient for accurate geospatial task programming, which requires coherent multi-step processes and multiple function calls. Compared to the baseline LLMs, the proposed GeoAgent has demonstrated superior performance, yielding notable improvements in function calls and task completion. In addition, these results offer valuable insights for the future development of LLM agents in automatic geospatial data analysis task programming.

Haoxiao Wang, Antao Xiang, Haiyang Sun et al.

Diffusion models are primarily trained for image synthesis, yet their denoising trajectories encode rich, spatially aligned visual priors. In this paper, we demonstrate that these priors can be utilized for text-conditioned semantic and open-vocabulary segmentation, and this approach can be generalized to various downstream tasks to make a general-purpose diffusion segmentation framework. Concretely, we introduce DiGSeg (Diffusion Models as a Generalist Segmentation Learner), which repurposes a pretrained diffusion model into a unified segmentation framework. Our approach encodes the input image and ground-truth mask into the latent space and concatenates them as conditioning signals for the diffusion U-Net. A parallel CLIP-aligned text pathway injects language features across multiple scales, enabling the model to align textual queries with evolving visual representations. This design transforms an off-the-shelf diffusion backbone into a universal interface that produces structured segmentation masks conditioned on both appearance and arbitrary text prompts. Extensive experiments demonstrate state-of-the-art performance on standard semantic segmentation benchmarks, as well as strong open-vocabulary generalization and cross-domain transfer to medical, remote sensing, and agricultural scenarios-without domain-specific architectural customization. These results indicate that modern diffusion backbones can serve as generalist segmentation learners rather than pure generators, narrowing the gap between visual generation and visual understanding.

Duochao Shi, Weijie Wang, Donny Y. Chen et al. · bytedance

Depth maps are widely used in feed-forward 3D Gaussian Splatting (3DGS) pipelines by unprojecting them into 3D point clouds for novel view synthesis. This approach offers advantages such as efficient training, the use of known camera poses, and accurate geometry estimation. However, depth discontinuities at object boundaries often lead to fragmented or sparse point clouds, degrading rendering quality -- a well-known limitation of depth-based representations. To tackle this issue, we introduce PM-Loss, a novel regularization loss based on a pointmap predicted by a pre-trained transformer. Although the pointmap itself may be less accurate than the depth map, it effectively enforces geometric smoothness, especially around object boundaries. With the improved depth map, our method significantly improves the feed-forward 3DGS across various architectures and scenes, delivering consistently better rendering results. Our project page: https://aim-uofa.github.io/PMLoss

Weijie Wang, Donny Y. Chen, Zeyu Zhang et al. · bytedance

Feed-forward 3D Gaussian Splatting (3DGS) models have recently emerged as a promising solution for novel view synthesis, enabling one-pass inference without the need for per-scene 3DGS optimization. However, their scalability is fundamentally constrained by the limited capacity of their models, leading to degraded performance or excessive memory consumption as the number of input views increases. In this work, we analyze feed-forward 3DGS frameworks through the lens of the Information Bottleneck principle and introduce ZPressor, a lightweight architecture-agnostic module that enables efficient compression of multi-view inputs into a compact latent state $Z$ that retains essential scene information while discarding redundancy. Concretely, ZPressor enables existing feed-forward 3DGS models to scale to over 100 input views at 480P resolution on an 80GB GPU, by partitioning the views into anchor and support sets and using cross attention to compress the information from the support views into anchor views, forming the compressed latent state $Z$. We show that integrating ZPressor into several state-of-the-art feed-forward 3DGS models consistently improves performance under moderate input views and enhances robustness under dense view settings on two large-scale benchmarks DL3DV-10K and RealEstate10K. The video results, code and trained models are available on our project page: https://lhmd.top/zpressor.

Haoxiao Wang, Kaichen Zhou, Binrui Gu et al.

Manipulating transparent objects presents significant challenges due to the complexities introduced by their reflection and refraction properties, which considerably hinder the accurate estimation of their 3D shapes. To address these challenges, we propose a single-view RGB-D-based depth completion framework, TransDiff, that leverages the Denoising Diffusion Probabilistic Models(DDPM) to achieve material-agnostic object grasping in desktop. Specifically, we leverage features extracted from RGB images, including semantic segmentation, edge maps, and normal maps, to condition the depth map generation process. Our method learns an iterative denoising process that transforms a random depth distribution into a depth map, guided by initially refined depth information, ensuring more accurate depth estimation in scenarios involving transparent objects. Additionally, we propose a novel training method to better align the noisy depth and RGB image features, which are used as conditions to refine depth estimation step by step. Finally, we utilized an improved inference process to accelerate the denoising procedure. Through comprehensive experimental validation, we demonstrate that our method significantly outperforms the baselines in both synthetic and real-world benchmarks with acceptable inference time. The demo of our method can be found on https://wang-haoxiao.github.io/TransDiff/

Weijie Wang, Wenqi Ren, Guofeng Mei et al.

State-of-the-art 3D point cloud registration methods rely on labeled 3D datasets for training, which limits their practical applications in real-world scenarios and often hinders generalization to unseen scenes. Leveraging the zero-shot capabilities of foundation models offers a promising solution to these challenges. In this paper, we introduce ZeroReg, a zero-shot registration approach that utilizes 2D foundation models to predict 3D correspondences. Specifically, ZeroReg adopts an object-to-point matching strategy, starting with object localization and semantic feature extraction from multi-view images using foundation models. In the object matching stage, semantic features help identify correspondences between objects across views. However, relying solely on semantic features can lead to ambiguity, especially in scenes with multiple instances of the same category. To address this, we construct scene graphs to capture spatial relationships among objects and apply a graph matching algorithm to these graphs to accurately identify matched objects. Finally, computing fine-grained point-level correspondences within matched object regions using algorithms like SuperGlue and LoFTR achieves robust point cloud registration. Evaluations on benchmarks such as 3DMatch, 3DLoMatch, and ScanNet demonstrate ZeroReg's competitive performance, highlighting its potential to advance point-cloud registration by integrating semantic features from foundation models.

Baojun Che, Yifan Chen, Zhenghao Huan et al.

This paper is concerned with the approximation of probability distributions known up to normalization constants, with a focus on Bayesian inference for large-scale inverse problems in scientific computing. In this context, key challenges include costly repeated evaluations of forward models, multimodality, and inaccessible gradients for the forward model. To address them, we develop a variational inference framework that combines Fisher-Rao natural gradient with specialized quadrature rules to enable derivative free updates of Gaussian mixture variational families. The resulting method, termed Derivative Free Gaussian Mixture Variational Inference (DF-GMVI), guarantees covariance positivity and affine invariance, offering a stable and efficient framework for approximating complex posterior distributions. The effectiveness of DF-GMVI is demonstrated through numerical experiments on challenging scenarios, including distributions with multiple modes, infinitely many modes, and curved modes in spaces with up to 100 dimensions. The method's practicality is further demonstrated in a large-scale application, where it successfully recovers the initial conditions of the Navier-Stokes equations from solution data at positive times.

Haoxuan Han, Weijie Wang, Zeyu Zhang et al.

Recent advancements in Vision-Language Models (VLMs) have significantly pushed the boundaries of Visual Question Answering (VQA).However,high-resolution details can sometimes become noise that leads to hallucinations or reasoning errors. In this paper,we propose Degradation-Driven Prompting (DDP), a novel framework that improves VQA performance by strategically reducing image fidelity to force models to focus on essential structural information. We evaluate DDP across two distinct tasks. Physical attributes targets images prone to human misjudgment, where DDP employs a combination of 80p downsampling, structural visual aids (white background masks and orthometric lines), and In-Context Learning (ICL) to calibrate the model's focus. Perceptual phenomena addresses various machine-susceptible visual anomalies and illusions, including Visual Anomaly (VA), Color (CI), Motion(MI),Gestalt (GI), Geometric (GSI), and Visual Illusions (VI).For this task, DDP integrates a task-classification stage with specialized tools such as blur masks and contrast enhancement alongside downsampling. Our experimental results demonstrate that less is more: by intentionally degrading visual inputs and providing targeted structural prompts, DDP enables VLMs to bypass distracting textures and achieve superior reasoning accuracy on challenging visual benchmarks.

Weijie Wang, Yeqing Chen, Zeyu Zhang et al.

Feed-forward 3D Gaussian Splatting (3DGS) has emerged as a highly effective solution for novel view synthesis. Existing methods predominantly rely on a pixel-aligned Gaussian prediction paradigm, where each 2D pixel is mapped to a 3D Gaussian. We rethink this widely adopted formulation and identify several inherent limitations: it renders the reconstructed 3D models heavily dependent on the number of input views, leads to view-biased density distributions, and introduces alignment errors, particularly when source views contain occlusions or low texture. To address these challenges, we introduce VolSplat, a new multi-view feed-forward paradigm that replaces pixel alignment with voxel-aligned Gaussians. By directly predicting Gaussians from a predicted 3D voxel grid, it overcomes pixel alignment's reliance on error-prone 2D feature matching, ensuring robust multi-view consistency. Furthermore, it enables adaptive control over Gaussian density based on 3D scene complexity, yielding more faithful Gaussian point clouds, improved geometric consistency, and enhanced novel-view rendering quality. Experiments on widely used benchmarks including RealEstate10K and ScanNet demonstrate that VolSplat achieves state-of-the-art performance while producing more plausible and view-consistent Gaussian reconstructions. In addition to superior results, our approach establishes a more scalable framework for feed-forward 3D reconstruction with denser and more robust representations, paving the way for further research in wider communities. The video results, code and trained models are available on our project page: https://lhmd.top/volsplat.

Zibo Xu, Qiang Li, Weizhi Nie et al.

Medical Visual Question Answering (MedVQA) aims to answer medical questions according to medical images. However, the complexity of medical data leads to confounders that are difficult to observe, so bias between images and questions is inevitable. Such cross-modal bias makes it challenging to infer medically meaningful answers. In this work, we propose a causal inference framework for the MedVQA task, which effectively eliminates the relative confounding effect between the image and the question to ensure the precision of the question-answering (QA) session. We are the first to introduce a novel causal graph structure that represents the interaction between visual and textual elements, explicitly capturing how different questions influence visual features. During optimization, we apply the mutual information to discover spurious correlations and propose a multi-variable resampling front-door adjustment method to eliminate the relative confounding effect, which aims to align features based on their true causal relevance to the question-answering task. In addition, we also introduce a prompt strategy that combines multiple prompt forms to improve the model's ability to understand complex medical data and answer accurately. Extensive experiments on three MedVQA datasets demonstrate that 1) our method significantly improves the accuracy of MedVQA, and 2) our method achieves true causal correlations in the face of complex medical data.

Chenxi Li, Weijie Wang, Qiang Li et al.

Text-driven object insertion in 3D scenes is an emerging task that enables intuitive scene editing through natural language. However, existing 2D editing-based methods often rely on spatial priors such as 2D masks or 3D bounding boxes, and they struggle to ensure consistency of the inserted object. These limitations hinder flexibility and scalability in real-world applications. In this paper, we propose FreeInsert, a novel framework that leverages foundation models including MLLMs, LGMs, and diffusion models to disentangle object generation from spatial placement. This enables unsupervised and flexible object insertion in 3D scenes without spatial priors. FreeInsert starts with an MLLM-based parser that extracts structured semantics, including object types, spatial relationships, and attachment regions, from user instructions. These semantics guide both the reconstruction of the inserted object for 3D consistency and the learning of its degrees of freedom. We leverage the spatial reasoning capabilities of MLLMs to initialize object pose and scale. A hierarchical, spatially aware refinement stage further integrates spatial semantics and MLLM-inferred priors to enhance placement. Finally, the appearance of the object is improved using the inserted-object image to enhance visual fidelity. Experimental results demonstrate that FreeInsert achieves semantically coherent, spatially precise, and visually realistic 3D insertions without relying on spatial priors, offering a user-friendly and flexible editing experience.

Weijie Wang, Guofeng Mei, Jian Zhang et al.

Point cloud registration approaches often fail when the overlap between point clouds is low due to noisy point correspondences. This work introduces a novel cross-attention mechanism tailored for Transformer-based architectures that tackles this problem, by fusing information from coordinates and features at the super-point level between point clouds. This formulation has remained unexplored primarily because it must guarantee rotation and translation invariance since point clouds reside in different and independent reference frames. We integrate the Gromov-Wasserstein distance into the cross-attention formulation to jointly compute distances between points across different point clouds and account for their geometric structure. By doing so, points from two distinct point clouds can attend to each other under arbitrary rigid transformations. At the point level, we also devise a self-attention mechanism that aggregates the local geometric structure information into point features for fine matching. Our formulation boosts the number of inlier correspondences, thereby yielding more precise registration results compared to state-of-the-art approaches. We have conducted an extensive evaluation on 3DMatch, 3DLoMatch, KITTI, and 3DCSR datasets.

Wenqi Ren, Weijie Wang, Meng Zheng et al.

Open-Set Domain Adaptation for Semantic Segmentation (OSDA-SS) presents a significant challenge, as it requires both domain adaptation for known classes and the distinction of unknowns. Existing methods attempt to address both tasks within a single unified stage. We question this design, as the annotation imbalance between known and unknown classes often leads to negative transfer of known classes and underfitting for unknowns. To overcome these issues, we propose SATS, a Separating-then-Adapting Training Strategy, which addresses OSDA-SS through two sequential steps: known/unknown separation and unknown-aware domain adaptation. By providing the model with more accurate and well-aligned unknown classes, our method ensures a balanced learning of discriminative features for both known and unknown classes, steering the model toward discovering truly unknown objects. Additionally, we present hard unknown exploration, an innovative data augmentation method that exposes the model to more challenging unknowns, strengthening its ability to capture more comprehensive understanding of target unknowns. We evaluate our method on public OSDA-SS benchmarks. Experimental results demonstrate that our method achieves a substantial advancement, with a +3.85% H-Score improvement for GTA5-to-Cityscapes and +18.64% for SYNTHIA-to-Cityscapes, outperforming previous state-of-the-art methods.

Weijie Wang, Jiagang Zhu, Zeyu Zhang et al.

We present DriveGen3D, a novel framework for generating high-quality and highly controllable dynamic 3D driving scenes that addresses critical limitations in existing methodologies. Current approaches to driving scene synthesis either suffer from prohibitive computational demands for extended temporal generation, focus exclusively on prolonged video synthesis without 3D representation, or restrict themselves to static single-scene reconstruction. Our work bridges this methodological gap by integrating accelerated long-term video generation with large-scale dynamic scene reconstruction through multimodal conditional control. DriveGen3D introduces a unified pipeline consisting of two specialized components: FastDrive-DiT, an efficient video diffusion transformer for high-resolution, temporally coherent video synthesis under text and Bird's-Eye-View (BEV) layout guidance; and FastRecon3D, a feed-forward reconstruction module that rapidly builds 3D Gaussian representations across time, ensuring spatial-temporal consistency. Together, these components enable real-time generation of extended driving videos (up to $424\times800$ at 12 FPS) and corresponding dynamic 3D scenes, achieving SSIM of 0.811 and PSNR of 22.84 on novel view synthesis, all while maintaining parameter efficiency.

Rihao Chang, He Jiao, Weizhi Nie et al.

Recent advances in large language models (LLMs) have enabled new possibilities in simulating complex physiological systems. We introduce Organ-Agents, a multi-agent framework that simulates human physiology via LLM-driven agents. Each Simulator models a specific system (e.g., cardiovascular, renal, immune). Training consists of supervised fine-tuning on system-specific time-series data, followed by reinforcement-guided coordination using dynamic reference selection and error correction. We curated data from 7,134 sepsis patients and 7,895 controls, generating high-resolution trajectories across 9 systems and 125 variables. Organ-Agents achieved high simulation accuracy on 4,509 held-out patients, with per-system MSEs <0.16 and robustness across SOFA-based severity strata. External validation on 22,689 ICU patients from two hospitals showed moderate degradation under distribution shifts with stable simulation. Organ-Agents faithfully reproduces critical multi-system events (e.g., hypotension, hyperlactatemia, hypoxemia) with coherent timing and phase progression. Evaluation by 15 critical care physicians confirmed realism and physiological plausibility (mean Likert ratings 3.9 and 3.7). Organ-Agents also enables counterfactual simulations under alternative sepsis treatment strategies, generating trajectories and APACHE II scores aligned with matched real-world patients. In downstream early warning tasks, classifiers trained on synthetic data showed minimal AUROC drops (<0.04), indicating preserved decision-relevant patterns. These results position Organ-Agents as a credible, interpretable, and generalizable digital twin for precision diagnosis, treatment simulation, and hypothesis testing in critical care.

Jin Wang, Weijie Wang, Boyuan Deng et al.

Traditional control and planning for robotic manipulation heavily rely on precise physical models and predefined action sequences. While effective in structured environments, such approaches often fail in real-world scenarios due to modeling inaccuracies and struggle to generalize to novel tasks. In contrast, humans intuitively interact with their surroundings, demonstrating remarkable adaptability, making efficient decisions through implicit physical understanding. In this work, we propose INTENTION, a novel framework enabling robots with learned interactive intuition and autonomous manipulation in diverse scenarios, by integrating Vision-Language Models (VLMs) based scene reasoning with interaction-driven memory. We introduce Memory Graph to record scenes from previous task interactions which embodies human-like understanding and decision-making about different tasks in real world. Meanwhile, we design an Intuitive Perceptor that extracts physical relations and affordances from visual scenes. Together, these components empower robots to infer appropriate interaction behaviors in new scenes without relying on repetitive instructions. Videos: https://robo-intention.github.io

Boyuan Deng, Luca Rossini, Jin Wang et al.

Adaptive recovery from fall incidents are essential skills for the practical deployment of wheeled-legged robots, which uniquely combine the agility of legs with the speed of wheels for rapid recovery. However, traditional methods relying on preplanned recovery motions, simplified dynamics or sparse rewards often fail to produce robust recovery policies. This paper presents a learning-based framework integrating Episode-based Dynamic Reward Shaping and curriculum learning, which dynamically balances exploration of diverse recovery maneuvers with precise posture refinement. An asymmetric actor-critic architecture accelerates training by leveraging privileged information in simulation, while noise-injected observations enhance robustness against uncertainties. We further demonstrate that synergistic wheel-leg coordination reduces joint torque consumption by 15.8% and 26.2% and improves stabilization through energy transfer mechanisms. Extensive evaluations on two distinct quadruped platforms achieve recovery success rates up to 99.1% and 97.8% without platform-specific tuning. The supplementary material is available at https://boyuandeng.github.io/L2R-WheelLegCoordination/

Weizhi Nie, Zichun Zhang, Weijie Wang et al.

Counterfactual medical image generation effectively addresses data scarcity and enhances the interpretability of medical images. However, due to the complex and diverse pathological features of medical images and the imbalanced class distribution in medical data, generating high-quality and diverse medical images from limited data is significantly challenging. Additionally, to fully leverage the information in limited data, such as anatomical structure information and generate more structurally stable medical images while avoiding distortion or inconsistency. In this paper, in order to enhance the clinical relevance of generated data and improve the interpretability of the model, we propose a novel medical image generation framework, which generates independent pathological and structural features based on causal disentanglement and utilizes text-guided modeling of pathological features to regulate the generation of counterfactual images. First, we achieve feature separation through causal disentanglement and analyze the interactions between features. Here, we introduce group supervision to ensure the independence of pathological and identity features. Second, we leverage a diffusion model guided by pathological findings to model pathological features, enabling the generation of diverse counterfactual images. Meanwhile, we enhance accuracy by leveraging a large language model to extract lesion severity and location from medical reports. Additionally, we improve the performance of the latent diffusion model on long-tailed categories through initial noise optimization.

Jin Wang, Rui Dai, Weijie Wang et al.

Enabling robots to autonomously perform hybrid motions in diverse environments can be beneficial for long-horizon tasks such as material handling, household chores, and work assistance. This requires extensive exploitation of intrinsic motion capabilities, extraction of affordances from rich environmental information, and planning of physical interaction behaviors. Despite recent progress has demonstrated impressive humanoid whole-body control abilities, they struggle to achieve versatility and adaptability for new tasks. In this work, we propose HYPERmotion, a framework that learns, selects and plans behaviors based on tasks in different scenarios. We combine reinforcement learning with whole-body optimization to generate motion for 38 actuated joints and create a motion library to store the learned skills. We apply the planning and reasoning features of the large language models (LLMs) to complex loco-manipulation tasks, constructing a hierarchical task graph that comprises a series of primitive behaviors to bridge lower-level execution with higher-level planning. By leveraging the interaction of distilled spatial geometry and 2D observation with a visual language model (VLM) to ground knowledge into a robotic morphology selector to choose appropriate actions in single- or dual-arm, legged or wheeled locomotion. Experiments in simulation and real-world show that learned motions can efficiently adapt to new tasks, demonstrating high autonomy from free-text commands in unstructured scenes. Videos and website: hy-motion.github.io/

Songxue Gao, Chuanqi Jiao, Ruidong Chen et al.

Point cloud completion aims to recover raw point clouds captured by scanners from partial observations caused by occlusion and limited view angles. This makes it hard to recover details because the global feature is unlikely to capture the full details of all missing parts. In this paper, we propose a novel approach to point cloud completion task called Point-PC, which uses a memory network to retrieve shape priors and designs a causal inference model to filter missing shape information as supplemental geometric information to aid point cloud completion. Specifically, we propose a memory operating mechanism where the complete shape features and the corresponding shapes are stored in the form of ``key-value'' pairs. To retrieve similar shapes from the partial input, we also apply a contrastive learning-based pre-training scheme to transfer the features of incomplete shapes into the domain of complete shape features. Experimental results on the ShapeNet-55, PCN, and KITTI datasets demonstrate that Point-PC outperforms the state-of-the-art methods.