Video diffusion alignment has been heavily relied on scalar rewards. These rewards are typically derived from learned reward models in human preference datasets, requiring additional training and extensive collection. Moreover, scalar rewards provide coarse, global supervision, offering limited prompt-generation mismatch credit assignment and making models prone to reward exploitation and unstable optimization. We propose Diffusion-DRF, a free, rich, and differentiable reward framework for video diffusion fine-tuning. Diffusion-DRF employs a frozen, off-the-shelf Vision-Language Model (VLM) as the critic, eliminating the need for reward model training. Instead of relying on a single scalar reward, it decomposes each user prompt into multi-dimensional questions with freeform dense VQA explanation queries, yielding information-rich feedback. By direct differentiable optimization over this rich feedback, Diffusion-DRF achieves stable reward-based tuning without preference datasets collection. Diffusion-DRF achieves significant gains both quantitatively and qualitatively, outperforming state-of-the-art Flow-GRPO by 4.74% in overall performance on unseen VBench-2.0.
Anticipating how users will interact with their surroundings is essential for proactive systems that provide timely assistance in extended reality and embodied applications. Prior work typically forecasts hand-object interactions directly from historical observations, leaving intention and contact geometry implicit. However, similar observation histories can precede semantically or geometrically distinct futures. Inspired by behavioral evidence that gaze can provide early cues for underlying intention, we propose Gaze2HOI, a gaze-aware framework that formulates hand-object interaction forecasting as a hierarchical problem comprising intention inference, contact grounding, and coupled hand-object motion generation. Specifically, Gaze2HOI first weights gaze cues by fixation strength and recency, and uses them to modulate the visual representations of a vision-language model (VLM) through Temporal Gaze Saliency Modulation (TGSM), thereby reducing semantic ambiguity in intention inference. Subsequently, an Anticipatory Contact Grounding (ACG) module maps these structured intentions into explicit contact, support, and placement fields, effectively mitigating geometric ambiguity. Driven by these explicit conditions, a diffusion-based model jointly generates the six-degree-of-freedom (6-DoF) object trajectories and future articulated motion of both hands. To evaluate Gaze2HOI, we further introduce EgoAIM, a phase-aware benchmark that organizes egocentric hand-object interactions into four phases defined by evolving hand-object relations and provides fine-grained intention annotations. Extensive experiments demonstrate that Gaze2HOI achieves state-of-the-art performance in both intention inference and motion forecasting, even approaching oracle-level end-to-end performance in specific interaction phases. Project page: https://gaze2hoi.github.io/.
Video generation has progressed from isolated clips to coherent short-form audiovisual sequences, and the bottleneck is shifting from rendering to semantic control. This report summarizes the observations and insights on multimodal understanding gained from successive generations of Kling. We argue that generation quality is the projection of multimodal understanding through the entire data--conditioning--alignment--evaluation chain, and we examine this chain module by module. Kling-MLLM, a shared multimodal foundation model that unifies image, video, audio, synchronized audio--video, and interleaved inputs, is specialized for each downstream module. Because a DiT is trained on far fewer training samples than a language model, we treat captioning as a semantic contract that maximizes description density, and extend it to reference-grounded captioning that makes reference bindings explicit. Prompt enhancement translates user intent into this caption distribution: basic PE is derived from captions of compelling videos and refined with text-level and generator-in-the-loop reinforcement learning, while agentic PE refines the story and retrieves compelling examples. We show that a reliable text encoder comes from perceptual training rather than from stronger language ability, and that candidates can be screened from text alone before end-to-end training. Reward modeling pairs a general reward model for explainable, localized critiques with a decision reward model for fast multidimensional scoring, treating perceptual bandwidth as a first-class design parameter. Finally, we review task-specific automatic evaluation, focusing on protocols that expose scoring shortcuts and help trace failures to their upstream source. We hope these lessons offer practical guidance for building the understanding stack behind controllable video generation.
Multimodal agents use web and image tools to answer questions beyond their internal knowledge. Some search tasks use images mainly to identify an initial entity, then rely on text retrieval. These tasks offer limited insight into how agents connect information across images, objects, and search steps. We therefore focus on visual context management. This involves organizing and using visual information and linking it to other findings during search. Visually intensive tasks require agents to combine object appearance, local details, or spatial relations with external information. We introduce VisSearch-Bench, with 204 human-written tasks across 7 main domains and 25 subdomains. It covers Set-wise Visual Search, Sequential Visual Search, Hybrid Wide-Deep Search, Competitive Visual Disambiguation, and Visual-Structure-Guided Search. Each task provides a reference answer, traceable evidence, and scoring rules. Accuracy measures complete task success, while Process Score measures the correctness and completeness of individual results and subquestion answers. Across 16 multimodal models, the best system fully solves only 26.47% of tasks, even with search tools. These results highlight the challenge of linking visual evidence with search results to produce complete answers.
Reconstructing three-dimensional (3D) coronary anatomy from sparse angiographic views could support percutaneous coronary intervention, but limited angular coverage and imperfect vessel segmentations introduce substantial geometric ambiguity. We propose a vessel-specific reconstruction framework combining a learned 3D diffusion prior, visual-hull initialization, and projection-based data consistency. The unconditional diffusion model is trained on coronary vessel masks without paired 2D--3D data and reused across different view counts without retraining. During inference, a scaled visual hull initializes reconstruction, while projection-based updates constrain diffusion refinement using patient-specific vessel observations. We evaluated the framework on simulated angiograms derived from ImageCAS computed tomography angiography volumes using two, four, and eight views at two contrast levels. Across 50 evaluation coronary-tree cases, the proposed method achieved higher mean Dice and clDice than visual hull, deep image prior (DIP), and hash-grid implicit neural representation (INR) at all view counts under reduced contrast. With eight views, improvements over DIP, the best-performing baseline at lower contrast, were approximately 0.13 in Dice and 0.22 in clDice. At higher contrast, performance remained competitive with hash-grid INR, the best-performing eight-view baseline. The preferred diffusion starting timestep depended on observation quality and view count, with excessive refinement reducing accuracy at higher contrast. Projection guidance improved both metrics across all conditions. Averaging stochastic reconstructions further improved accuracy, while sample disagreement localized errors with average-precision gains of 0.14--0.21 over error prevalence. These findings support diffusion-prior reconstruction from sparse vessel observations, although clinical validation remains necessary.
Jimmy Wang, Yifan Chou, Kaifeng Zhang, Yixuan Wang, Binghao Huang, Tao Yu, Saleh Nabi, Yunzhu Li
Predicting how the world evolves in response to robot actions is central to policy learning, evaluation, and planning. However, many action-conditioned world models predict future observations in 2D image or video space without explicitly modeling the underlying 3D scene, making spatial consistency across viewpoints and over long-horizon interactions challenging. We introduce 3D World Simulator, an action-conditioned multimodal world modeling framework that jointly predicts appearance, geometry, and tactile observations from a unified 3D representation. Our approach combines a hybrid particle-voxel autoencoder that compresses point clouds into compact 3D voxel latents with an action-conditioned latent dynamics model that predicts their evolution. The shared 3D representation supports geometrically consistent rendering from arbitrary viewpoints and flexible integration of spatially aligned tactile signals. Our experiments demonstrate interactive, long-horizon, multimodal prediction at 50 Hz. We then show how our model enables scalable data generation for policy training and evaluation using tactile and multi-view visual observations, including wrist-mounted and novel camera views. Furthermore, we scale our method to large datasets containing diverse objects and demonstrate model-based planning with unseen objects and robot configurations. Together, these results highlight the potential of hybrid particle-voxel dynamics as a foundation for multimodal 3D world simulation.
Moving Object Segmentation (MOS) detects and separates dynamic objects from the static background in a video, which plays a fundamental role in dynamic-scene understanding. Most MOS methods rely on pure visual motion cues alone, such as optical flow or point tracks, which is inherently ambiguous due to the compounded camera and object motions as well as motion-induced appearance changes, e.g. shadows and reflections. Naive fusion of additional information such as depth and semantics, however, cannot leverage their complementarity to the fullest extent. We therefore propose DynOSeg, a novel MOS approach that takes on a two-stage fusion structure. DynOSeg first combines a geometry-aware residual map and depth encoding into a geometric representation of motion, delineating camera motions and object dynamics. We then further incorporate semantic features to refine object understanding. We have additionally curated a diverse training corpus including CARLA-MOS, a new synthetic urban-driving dataset for MOS training. Our experiments show that DynOSeg consistently outperforms published state-of-the-art MOS methods across a diverse range of datasets. Our code and pretrained models can be found at https://github.com/ACFR-RPG/DynOSeg
The edge detection technique is an essential step for aircraft skin milling in aviation manufacturing. Most of the current detection methods focus on traditionally defined edge extraction tasks but disregard the crucial systematic requirement of edge milling. In this paper, we proposed a novel edge detection framework for automatic edge milling of aircraft skins. First, an edge probability detector is proposed by the spatial tangent continuity (STC) to provide the essential reference. Second, we propose a hierarchical branch searching (HBS) method to hierarchically strip the desired milling edges from the raw point cloud, which consists of three graded progressive steps: branch backbone generation, branch extension, and branch pruning. We demonstrate the performance of the proposed method on both synthetic models and aircraft skin workpieces. The proposed method outperforms the other baselines and shows accurate edges for the edge milling task.
The advancement of multimodal learning is heavily constrained by the scarcity and high annotation cost of high-quality instruction data. While raw videos offer an abundant and dynamic source of visual knowledge, transforming them into precise instruction data remains highly challenging. To bridge this gap, we present VidUEU-Agent, an automated pipeline that synthesizes instruction data from raw videos for multimodal understanding, editing, and unified tasks. VidUEU-Agent follows a four-stage pipeline: it performs global perception to filter low-quality shots, mines keyframes and multimodal signals, constructs metadata and routes the metadata to synthesize instruction samples. More importantly, our agent framework can autonomously iterate the data distribution based on downstream training results, thereby optimizing data quality, balance, and training effectiveness. Scaling this pipeline, we construct VidUEU-100K, a large-scale dataset of 100K training samples spanning multiple domains and tasks, providing a robust foundation for developing comprehensive multimodal capabilities. Extensive experiments demonstrate that fine-tuning baseline models on VidUEU-100K yields significant and consistent performance gains across diverse representative benchmarks, validating the effectiveness of our proposed agent framework and dataset.
Huihan Wang, Juncheng Wang, Yuxiang Feng, Wenlong Hou, Guoqi Yu, Chao Xu, Yang Liu, Baigui Sun 等 9 位作者
Missing modality imputation aims to synthesize unobserved modalities from observed ones. Existing methods treat each modality as living on its own latent manifold, forcing the generator to learn a long, unstructured jump between them. We present SphereFlow, which takes a fundamentally different view: by projecting all modalities into a shared hyperspherical latent space via a frozen self-supervised encoder, cross-modal imputation reduces to short-range geometric transport between nearby points on the sphere. Realizing this idea with flow matching, however, exposes two geometric conflicts: the standard Gaussian noise source is catastrophically mismatched with the unit-norm data manifold, and the linear interpolation path departs the sphere surface into a region where the decoder has no training signal. We resolve both with simple, geometry-aware modifications: replacing the noise source with the observed-modality latent so that the velocity field learns only the cross-modal displacement, and training the decoder with a time-weighted loss that extends its domain to the near-surface region traversed by the flow. We prove that this data-as-source formulation reduces the worst-case off-manifold deviation from a quantity that grows with the latent dimension to a small, dimension-independent bound governed by inter-modality similarity. Experiments on BraTS multi-modal MRI imputation and CT-MRI translation demonstrate state-of-the-art generation fidelity across all settings, while downstream brain tumor segmentation with our imputed data closes roughly eighty percent of the gap to the full-modality oracle. Code is available at https://github.com/WTHH031230/SphereFlow.
Zhongquan Wang, Long Cao, Peijia Xie, Senlei Ma, Jie Li, Guofa Li
High-fidelity weather editing of 4D scenes can support end-to-end training and closed-loop simulation for autonomous driving. Combining generative image editing with dynamic Gaussian representations enables multi-weather scene reconstruction, but weather-edited images often contain structural distortions. Directly fitting their appearance may propagate errors from teacher images into the reconstructed 4D scene, compromising the original scene structure and temporal continuity. To address these challenges, we propose IGSW, a two-stage framework that combines intrinsic-decoupled 4D Gaussian scene modeling with structure-preserving weather refinement. In Stage I, background Gaussian appearance is decomposed into weather-shared albedo and normals and weather-specific irradiance and residuals. Weather teacher images generated by a diffusion-based image editing model and intrinsic maps estimated by a diffusion-based inverse renderer provide feature-level soft supervision for stable structure and illumination changes. In Stage II, a structure-preservation gate combines semantic conflicts with cross-weather intrinsic discrepancies to suppress unreliable teacher regions. Per-Gaussian weather transport variables and target-weather branches then strengthen the desired weather appearance over the intrinsic-decoupled scene. Experiments on nuScenes and Waymo show that IGSW achieves the highest content similarity and semantic consistency and the lowest temporal error among the compared methods, while retaining pronounced weather-dependent illumination and atmosphere. The project page is available at https://akaxi6.github.io/IGSW/ .
Photorealistic Style Transfer (PST) aims to transfer the color and tonal style of a reference to a content image while strictly preserving its structural integrity. However, existing deep learning-based methods inherently suffer from semantic entanglement caused by pre-trained image encoders, leading to unnatural spatial distortions. Moreover, current pixel-level mapping paradigms often ignore color gamut topology, resulting in color banding, while also lacking the multimodal capability for precise, tone-specific text-driven control. To address these bottlenecks, we propose StatLUT, an innovative multimodal framework for 3D LUT generation. First, we bypass traditional encoders and introduce a Lab-Extractor to derive spatially-agnostic statistical features, fundamentally decoupling color distributions from structural semantics to ensure artifact-free rendering. Second, we formulate LUT generation as a Transformer-based Seq2Seq translation task, utilizing a Multi-dimensional Residual Mapper (MR-Mapper) to predict topologically smooth 3D LUTs. Finally, to break the unimodal barrier, we propose the H-Diffuser, a lightweight Diffusion Transformer that directly synthesizes statistical features from tone-descriptive text prompts, enabling controllable and robust color grading. Extensive experiments on standard benchmarks demonstrate that StatLUT significantly outperforms state-of-the-art methods in both visual quality and quantitative metrics, pioneering a highly robust and versatile paradigm for multimodal photorealistic style transfer.
Tengfei Liu, Yang Shi, Yuran Wang, Yuqi Tang, Yizhuo Jia, Zhuoran Zhang, Qixun Wang, Shujie Li 等 15 位作者
Multimodal large language models (MLLMs) can generate detailed video captions, yet they may overemphasize minor details while overlooking essential context and key events. We argue that not all visual evidence deserves equal attention, as the importance of visual information depends on the content of each video. Motivated by this observation, we propose Reviso, an agentic video captioning framework that learns what to observe and how to gather relevant visual evidence. From a low-frame-rate global preview, Reviso constructs a hierarchical Context--Focus--Points (CFP) observation plan that identifies the video context, prioritizes key aspects, and defines concrete observation targets and corresponding tools. Guided by this plan, Reviso performs goal-directed active perception by selectively revisiting temporal segments and combining tools for fine-grained inspection. It interprets the acquired evidence, updates its plan as needed, and integrates global and local information to generate the final caption. To learn this observation strategy, we develop a verification-guided pipeline that constructs high-quality training trajectories through factual verification and teacher-guided refinement. We further train Reviso using Group Relative Policy Optimization (GRPO), incorporating a key-fact reward for caption quality, a Plan Alignment Reward (PAR) for identifying important observation targets, and an Unproductive Tool-Call Penalty (UTCP) that adjusts observation-cost penalties based on caption quality. Experiments across four video captioning benchmarks show that Reviso outperforms the evaluated open-source baselines, while its generated captions better support downstream video question answering. These findings highlight the effectiveness and potential of priority-guided active observation for video captioning.
Video dubbing aims to synthesize lip movements that match target speech while preserving the speaker identity, head pose, and background of a source video. Existing editors are predominantly trained via self-reconstruction under original paired audio, even though the source mouth becomes a conflicting target when driving speech is replaced. We introduce replacement-aware cross-audio supervision: paired-audio samples use full-frame reconstruction, whereas unpaired replacement-audio samples mask out source-mouth reconstruction losses and supervise the mouth solely through audio-visual synchronization while preserving unmasked facial regions and background. Because offline dubbing provides the complete source video and target speech beforehand, we present VidDub, a full-context diffusion editor distilled into two denoising steps (2 NFE). Our replacement-aware teacher first learns accurate lip synchronization without regressing toward conflicting source mouth movements. Matched trajectory regression then transfers the teacher's CFG-guided trajectories to a two-step student, and distribution matching distillation recovers visual quality after step compression. We will release the curated training dataset, VidDub-Bench, fixed paired- and cross-audio evaluation manifests, preprocessing outputs, and unified evaluation code. Across three training seeds, masking out conflicting source-mouth reconstruction improves cross-audio Sync-C by 0.833; the final two-step model combines strong lip synchronization with superior visual quality, identity preservation, and visual speech recognition across four benchmarks.
World models (WMs) have demonstrated strong potential for end-to-end autonomous driving by learning predictive representations of future scene dynamics. However, generating future videos during inference introduces substantial computational overhead, leading many recent driving WMs to adopt a single front camera as input for efficient deployment. This design restricts spatial coverage in safety-critical maneuvers such as lane changes, merges, and turns. To address this limitation, we propose SV-WAM, a surround-view world-action model (WAM) that preserves full six-camera observations while maintaining efficient inference. SV-WAM leverages future-video prediction as dense training supervision for action learning within a shared generative model, rather than as an inference-time output. At the core of this design is an action-centered causal mask that prevents action tokens from attending to future-video tokens during joint action-video denoising. Consequently, the video branch can be discarded at deployment, enabling efficient action-only planning. Furthermore, we introduce a differentiable drivable-area compliance regularizer that penalizes vehicle-footprint corners approaching or crossing drivable boundaries, improving planning safety and boundary awareness. Extensive experiments on the closed-loop NAVSIMv2 benchmark and the open-loop nuScenes benchmark demonstrate that SV-WAM achieves state-of-the-art planning performance with low inference latency and competitive zero-shot transfer capability.
Xinyue Xu, Hongbin Lin, Juangui Xu, Hualiang Wang, Lehan Wang, Lijie Hu, Adrian Weller, Xiaomeng Li
Medical imaging modalities such as ultrasound and X-ray are widely used in clinical practice, where diagnosis follows a structured, evidence-driven workflow aligned with standardized criteria. While multimodal large language models (MLLMs) show promise for automated medical report generation, most existing systems rely on end-to-end multimodal fusion without modeling clinically defined intermediate attributes, leading to limited grounding and interpretability. To address this issue, we propose CORAL (COncept-grounded ReAsoning with Localization), a multimodal framework that integrates spatial grounding and concept-level supervision into a unified reasoning process. CORAL employs a prompt-driven medical segmentation model to localize lesions and predicts multi-class clinical attributes through a Concept Bottleneck module. The resulting textual concept tokens are combined with mask-modulated visual features within an MLLM to enable structured report generation and diagnostic prediction. Experiments on BUS-CoT and IU X-ray datasets demonstrate consistent improvements in diagnostic accuracy, concept consistency, and report quality over strong general-purpose and medical MLLMs, indicating that concept-grounded reasoning better aligns generation with clinical decision processes.
Tiansheng Hu, Yichuan Wang, James V. Talwar, Nicholas Tomlin, Chen Zhao, Sewon Min
Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existing deep research (DR) agents typically reduce papers to OCR-extracted text. We instead ask whether vision-language models (VLMs) can enable page-native deep research: directly retrieving and reasoning over rendered document page images. We introduce DR Pixel, the first open-source end-to-end multimodal DR agent that uses rendered page images as its sole retrieval and reasoning unit, operating over a corpus of 1.06M pages. To evaluate this setting, we construct DRPixelBench, a benchmark on scientific papers requiring reasoning over text, figures, tables, and equations. Surprisingly, page-native DR already works well without training, outperforming OCR-based baselines across state-of-the-art models like GPT-5.4, Gemini-3.5-Flash, and Qwen3.6-27B, while requiring fewer tool calls. We next turn to DR training. A key challenge is that teacher trajectories for agentic supervised finetuning are noisy, mixing critical reasoning and action steps with redundant or failed attempts. To address this, we introduce counterfactual credit reweighting, a fine-grained credit assignment method that weights individual reasoning and action steps by their counterfactual contribution to the final answer. Our training improves retrieval recall by about 15 points and final accuracy by over 10 points on DRPixelBench, and likewise outperforms the untrained baseline on four existing benchmarks. On out-of-distribution ViDoRe-V3, DR Pixel comes within 1.8 points of an oracle that is given the gold pages directly. Our results suggest that the conventional OCR-then-retrieve pipeline is not only unnecessary but suboptimal: directly retrieving and reasoning over visual document pages makes complex DR both more effective and more efficient.
Marcos Antonio Obando, Tianyuan Wang, Tanja Tarvainen, Tristan van Leeuwen, Felix Lucka
Adaptive compressed sensing (CS) aims to select the most informative measurements sequentially, which is crucial for many modern imaging systems. This task can be formulated as a Bayesian sequential optimal experimental design (sOED) problem, where each new acquisition is chosen to maximally reduce uncertainty about the unknown image given prior information and subsequent data provided by acquisitions. However, in single-pixel imaging (SPI), the large design space consists of all possible binary illumination patterns. Directly optimizing the Bayesian sOED objective over this space using conventional methods is generally challenging. In this work, we reformulate sOED for SPI as a reinforcement learning (RL) problem. In RL, an agent learns a policy through repeated experimental trials to maximize a reward, here given as the expected information gain, the most complete measure of uncertainty reduction in Bayesian OED. This avoids the need to directly optimize complex information-theoretic objectives during a measurement to obtain the next design. Instead, the result is a learned policy function that directly maps the current posterior belief about the image to the next optimal binary measurement pattern, in a computationally efficient way. Using a simple yet representative SPI imaging scenario, this work evaluates the scalability of the learned policy under two different policy parameterizations. Furthermore, we investigate how performance deteriorates when only approximate posterior information is available or the measurements are impaired by noise, and propose several strategies to introduce additional design constraints motivated by biomedical applications. The numerical experiments demonstrate that the proposed Bayesian sOED-RL framework remains robust and can be extended to more complex, large-scale SPI scenarios.
Adapting features before upsampling can improve dense prediction, but the gains may arise from changes in feature values, spatial aggregation, or both. We investigate this ambiguity by comparing matched residual adapters before and after a frozen JAFAR upsampler under a shared parent. We construct a readout-null intervention that preserves a frozen linear classifier's coarse logits while allowing spatial attention weights to adapt. Guidance/value exchanges on trained models and retrained fixed-guidance controls test the utility of this adaptation route. In three-seed experiments on PASCAL Visual Object Classes (VOC) and Cityscapes, adaptation before upsampling improves mean intersection over union (mIoU) by 0.807 and 1.413 percentage points, respectively, over adaptation after upsampling. Across both datasets and the JAFAR and AnyUp upsamplers, readout-null adaptation produces mean gains of 1.379-1.987 points over the corresponding frozen linear baselines. Retrained controls further support the benefit of adaptive spatial guidance. These findings explain how dense predictions can improve despite unchanged coarse classifier outputs and identify spatial guidance as a useful target when adapting frozen feature upsampling systems.
Dayou Li, Hao Wang, Qianqian Yang, Zihao Zhu, Haoquan Fang, Ziyao Zeng, Yan Han, Zihan Wang 等 26 位作者
Large-scale egocentric human interaction data is becoming an important source of physical supervision for embodied learning, yet video alone leaves the contact and pressure that characterize physical interaction unrecorded. Recent visual-tactile datasets provide this missing supervision, but their synchronized tactile data remain far smaller in volume than human video. Moreover, the largest resources often merge recordings from different sensors or annotation procedures, which makes the effect of data scale difficult to isolate. We therefore introduce TouchScale, a 500-hour dataset of contact-rich human interaction recorded with a single unified wearable setup. Its approximately 2K predefined task descriptions span everyday activities and structured manipulation, and each recording temporally aligns egocentric RGB-D video with wrist RGB video and dense full-hand bimanual tactile measurements. Compared with prior tactile data, training on the full TouchScale raises zero-shot contact IoU on data from an unseen tactile sensor from 0.134 to 0.383. Pretraining a visual encoder on TouchScale also yields the highest action recognition accuracy on three benchmarks among the compared visual-tactile datasets. Used for visual-tactile mid-training of a robot policy, TouchScale improves the average real-world success rate across four contact-rich manipulation tasks from 22.5% to 57.5%. With the sensor and collection protocol held fixed, both zero-shot tactile prediction and robot success show an overall upward trend as more TouchScale data is used. These results suggest that human visual-tactile data collected at scale with consistent sensing benefits both perception and robot manipulation. We will publicly release TouchScale, including all synchronized visual-tactile recordings and reconstructed object models, to support future research on scalable visual-tactile learning.
Haoyang Yuan, Boyang Li, Yingqian Wang, Yimian Dai, Nuo Chen, Xinfei Huang, Shuqi Yi, Zaiping Lin 等 10 位作者
Omni-domain infrared small target (IRST) detection is crucial for infrared surveillance, yet remains challenging due to heterogeneous imaging domains and inconsistent target characteristics. Previous deep learning-based methods have been developed for visual-only paradigms and achieved promising performance on domain-specific tasks. However, existing methods follow the task-specific supervised learning paradigm. This paradigm simplifies the full-scene infrared observations to sparse target supervision, discarding the semantics that remain invariant across heterogeneous domains. Consequently, detection performance suffers substantially under domain shifts. To handle this issue, we introduce \textbf{``understand before detect''}, a paradigm that formulates omni-domain IRST detection as an understanding-driven process, where holistic infrared target understanding precedes precise detection. Building on this paradigm, we propose \textbf{JinSight}, which first develops holistic IRST understanding through language supervision and then transfers the learned cross-domain representations to precise small-target detection. By grounding infrared representations in language semantics, JinSight enables a single model to generalize across heterogeneous infrared domains. We then introduce Latent Semantic Interaction (LSI), which exchanges language-aligned global semantics with fine-grained spatial features in a compact low-rank space. To address the lack of multimodal omni-domain IRST benchmarks, we build \textbf{OmniIRST-VL}, the first large-scale, highly diverse vision--language dataset for omni-domain IRST detection. It comprises over 39k annotations across six complementary instruction tasks covering both scene-level understanding and target-centric reasoning.
Manni Cui, Ziheng Qin, ZiAn Wang, Ruiqi Liu, Dianyuan Zou, Jianglan Wei, Han Zhou, Yu Liu 等 11 位作者
AI-generated videos (AIGVs) typically contain subtle temporal artifacts that arise from inter-frame inconsistencies rather than within individual frames. A detector that captures such artifacts should therefore benefit from video pretrained backbones over image only ones. In practice, however, video backbones with standard global readouts often fail to outperform strong image pretrained probes on AIGV benchmarks. We attribute this gap to excessive spatiotemporal aggregation in the readout. Video pretrained backbones tend to compress each frame into a single global descriptor. This compression suppresses local patch level temporal dynamics and discards inter patch relations, which are precisely the cues that AIGV detection most reliably depends on. Based on this, we propose Velocity Gated Patch Velocity Profiling (V-PVP), a lightweight readout that replaces only the aggregation layer with two parallel streams over the patch velocity field, adding only about $0.5$M trainable parameters. V-PVP serves as a general plug-and-play module that consistently improves performance across diverse video backbones under both end-to-end fine-tuning and linear probing settings. Our method reaches 95.28 AUC on AIGVDBench while keeping the backbone fully frozen. The results show that simply replacing the aggregation layer reactivates the temporal potential of frozen video backbones, restoring their advantage on AIGV detection. Code is available at https://github.com/manic-cui/Rethinking-the-readout.
Shuiwen Shen, Jinhua Zhang, Yan Wang, Huanqiang Zeng, Yueyun Xu, Song Gao, Haifeng Xiang, Yifu Shen
The transition of information-centric artificial intelligence to safety-critical applications is impeded by the limitations of conventional offline-training paradigms in open-world environments. This paper presents Control AI, an architecture for Online Active Learning (OAL) inspired by the continuous Prediction-Evaluation-Correction process of biological intelligence. The framework combines the hierarchical representation capacity of deep neural networks with the real-time error-feedback adaptation of adaptive control. A prediction model maps sensory input into a latent manifold, an optional latent model represents transitions in that space, and a reconstruction model maps the latent state back to observable signals. Residuals between predictions and measurements provide the internal learning signal. Spatial and temporal cross-reconstruction architectures prevent OAL from collapsing to a trivial identity mapping, while coupled, dual, and localized variants shorten error-transmission paths. Localized predictive coding style OAL is implemented through guidance-backpropagation and target-backpropagation architectures, and hybrid and scaled variants support high-dimensional deployment. The paper focuses on the architectural rationale and the progression from a conceptual feedback loop to practical large-scale configurations.
Multimodal large language models (MLLMs) hold immense potential to revolutionize clinical practice, yet deploying them in the medical domain is fundamentally a vision-centric challenge: models must absorb knowledge from heterogeneous 2D and 3D medical images, and evaluation protocols must align with radiologists' clinical practice and provide an accurate, fine-grained and factualness-driven assessment. In this paper, we introduce ClinFusion, a vision-centric MLLM designed for holistic medical understanding that systematically addresses these limitations. We propose a compositional and cascaded vision encoder architecture featuring a Cascade Spatial-Aware Locality Fusion operator that unifies diverse 2D and native 3D medical image understanding within a fused encoder. We further introduce a vision-grounded evaluation framework, including MedIF-Bench for instruction-following assessment and a region-of-interest-grounded method for clinically aligned and factualness-driven report generation evaluation. We show that ClinFusion sets a new state-of-the-art across a comprehensive suite of 2D and 3D multimodal medical benchmarks---spanning visual question answering, report generation, and instruction following---as well as textual medical tasks, outperforming leading open-source medical MLLMs (\textit{e.g.}, Hulu-Med, Lingshu) on 20 out of 24 benchmarks and demonstrating multimodal capabilities better than powerful proprietary models such as GPT-5.2 and Gemini-3-Flash on 13 out of 16 benchmarks, and can be further augmented with agentic tool use for retrieval-augmented and tool-assisted clinical workflows. A blinded evaluation by board-certified radiologists confirms that ClinFusion produces the highest-ranked reports, and validates our RoI-grounded metric as achieving the strongest correlation with expert judgment among all automatic evaluation metrics examined.
Yongqi Shan, Yifan Wang, Yunzhi Zhuge, Lijun Wang, Huchuan Lu
Understanding and reasoning over 3D scenes from point clouds is essential for a broad range of AI applications. Recent 3D large language models have shown promising results, yet they still face two fundamental limitations: (i) insufficient capability for grounded multi-step reasoning in complex 3D scenes, and (ii) a text-centric inference paradigm that lacks effective interaction between textual reasoning and visual evidence. As a result, these models often fail to verify intermediate inferences, disambiguate similar object instances, and avoid hallucinations. To address these limitations, we propose MMCoT$^3$, the first 3D-Text Interleaved Chain-of-Thought (CoT) framework for 3D scene understanding. The core idea is to explicitly interleave textual reasoning with task-relevant 3D evidence, enabling the model to consult scene observations throughout the reasoning process. In addition, since a textual mention in 3D scenes may correspond to multiple candidate objects, naive object-mention alignment can inject incorrect evidence and mislead subsequent reasoning. To mitigate this multi-instance ambiguity, we design an ambiguity-aware evidence aggregation mechanism that uses query-guided cross-attention to fuse candidate object tokens and inject both evidence and ambiguity cues into the reasoning process. To better equip the model with multi-step reasoning capability, we further develop PointCoT, a large-scale 3D-specific CoT instruction tuning dataset with approximately 176.2K instances spanning captioning, question answering, dialogue, and planning tasks. Extensive experiments on six benchmarks show that our method consistently improves performance and achieves state-of-the-art results, validating the effectiveness of our proposed ambiguity-aware 3D-text interleaved reasoning for 3D scene understanding.
Quantitative phase imaging (QPI) reveals optical path-length variations in transparent objects but typically relies on spatially resolved intensity measurements. Here we numerically and experimentally demonstrate learned optical compression for single-pixel QPI, jointly optimizing a diffractive encoder and a digital neural decoder to concentrate reconstruction-relevant phase information into a compact measurement vector. A single reconfigurable phase-modulation plane sequentially displays the learned encoding patterns, which are independent of the input object, each yielding one scalar intensity measurement. Operating at a visible wavelength, the experimental system reconstructs quantitative phase images of unknown objects with ~40-fold dimensional compression, using only 2.55% as many intensity measurements as reconstructed pixels. Experimental blind testing on handwritten digits and grating-like phase objects achieves mean Pearson correlation coefficients of 0.945 and 0.848, respectively. Benchmarking under the same measurement budget demonstrates substantially higher reconstruction fidelity than other encoding strategies, highlighting the advantages of learned optical compression. With the optical encoder fixed, decoder-only neural adaptation enables experimental transfer to previously unseen data distributions, extending the utility of the learned measurement representation. These results establish learned optical compression as a measurement-efficient and reusable framework for QPI, integrating phase-information compression into single-pixel intensity measurements while retaining adaptability to new data. This approach paves the way for low-data-rate phase imaging and sensing, with the potential to extend QPI to spectral regions where high-quality detector arrays are costly or unavailable.
Jianfeng Ding, Linfeng Tang, Hao Zhang, Han Xu, Xunpeng Yi, Yibing Zhang, Meiqi Gong, Zhongyuan Wang 等 9 位作者
In the absence of ground truth, most existing infrared and visible image fusion methods pursue fidelity to the source images and directly superimpose two physically distinct modalities. As a result, salient infrared targets appear as bright patches of thermal radiation pasted onto the scene. These patches obscure the original color and texture of the targets, so the fused images look unnatural to human observers and deviate from the data distribution learned by vision models pretrained on visible images. Therefore, we redefine infrared and visible image fusion as natural complementary enhancement rather than source fidelity, and propose NatureFusion, which exploits cross-modal complementary cues to infer a clear and natural visible-style image from low-quality visible and infrared image pairs. Specifically, to address the lack of ground truth, we employ a generative foundation model as the teacher to directly generate visually natural fused images for supervision. Furthermore, to fully distill the teacher's generative capability while reducing computational cost, we fine-tune a pretrained generative model with conditional flow matching through a light-weight fusion adapter. Last but not least, to further improve downstream performance without sacrificing natural appearance, we freeze the fusion adapter and introduce a detector-supervised task adapter that strengthens the complementary contribution of infrared. Extensive comparisons with seven multi-modal fusion and seven single-modality enhancement methods demonstrate that NatureFusion generates visually natural fused images with the best perceptual quality. Detection and segmentation models pretrained on visible images achieve substantially higher accuracy on our fused images than on those of any competing method.
Building general game agents requires both high-level task reasoning and low-latency action control in rapidly changing visual environments. These capabilities favor models with substantially different computational profiles, making them difficult to realize within a single model. In this work, we present TaskFrame, a hierarchical planner-controller framework that connects task-level planning with frame-level keyboard-and-mouse control. A vision-language task planner decomposes a high-level goal into an ordered sequence of atomic instructions and is optimized through supervised learning and dependency-aware GRPO. The hidden states of special goal tokens that terminate each atomic instruction provide continuous goal embeddings that are cached and reused across control steps. A causal Transformer controller combines the embedding of the active goal with recent observations to predict discrete keyboard and mouse-button actions and continuous camera motion at every game step. We train the system in three stages: planner training, controller pretraining, and joint alignment. Experiments in Minecraft show that TaskFrame is competitive with general-purpose vision-language models on instruction decomposition, and that in closed-loop evaluation the agent achieves a higher combat success rate than the evaluated baselines.
Minyan Luo, Yuxin Zhang, Meng Zhao, Xincan Wang, Fuzhang Wu, Tong-Yee Lee, Oliver Deussen, Fan Tang 等 9 位作者
An outstanding creative advertising image should both attract attention and communicate a product's positioning. How to choose a persuasive conceptual bridge before rendering begins is challenging for advertising generation. We present AdMuse, a skill-guided multi-agent framework that distills high-quality advertisements into reusable creative skills for cross-domain advertising strategy. Rather than retrieving isolated exemplars, AdMuse abstracts recurring bridge patterns into a skill library that can guide new product briefs. At generation time, AdMuse coordinates skill routing, multi-candidate ideation and selection, and VLM-guided visual refinement: diverse skills propose alternative concepts, relative judgment selects the most promising direction, and vision-language feedback helps realize the selected intent in the final image. This design treats advertising creativity as a structured decision process rather than a single prompt-writing step. Experiments across diverse product and brand positionings show that AdMuse produces visually compelling advertisements that faithfully communicate product positioning.
In this paper, we propose Concentrate and Concentrate (CaC), a coarse-to-fine anomaly reward model based on Vision-Language Models. During inference, it first conducts a global temporal scan to anchor anomalous time windows, then performs fine-grained spatial grounding within the localized interval, and finally derives robust judgments via structured spatiotemporal Chain-of-Thought reasoning. To equip the model with these capabilities, we construct the first large-scale generated video anomaly dataset with per-frame bounding-box annotations, temporal anomaly windows, and fine-grained attribution labels. Building on this dataset, we design a three-stage progressive training paradigm. The model initially learns spatial and temporal anchoring through single- and multi-frame supervised fine-tuning, and then is optimized by a reinforcement learning strategy based on two-turn Group Relative Policy Optimization (GRPO). Beyond conventional accuracy rewards, we introduce Temporal and Spatial IoU rewards to supervise the intermediate localization process, effectively guiding the model toward more grounded and interpretable spatiotemporal reasoning. Extensive experiments demonstrate that CaC can stably concentrate on subtle anomalies, achieving a 25.7% accuracy improvement on fine-grained anomaly benchmarks and, when used as a reward signal, CaC reduces generated-video anomalies by 11.7% while improving overall video quality.