多模态信息融合

Existing multi-speaker dialogue systems bind speakers to utterances through structured supervision: per-turn tags, multi-stream transcriptions, or learnable speaker embeddings. These systems operate within speech-only pipelines that produce clean vocal sequences without the ambient texture of real conversations. We take a different approach. Our method, ScenA, conditions a text-to-audio flow-matching foundation model, pretrained on large-scale in-the-wild data, directly on multiple reference voices and a free-form natural language prompt that describes an entire multi-speaker audio scene. Leveraging such a foundational model allows us to inherit its capacity for natural, non-studio audio: background noise, room acoustics, overlapping dialogue, and spontaneous paralinguistic events, while adding multi-speaker control without any per-turn structure. Concretely, reference latents are concatenated into the model's token sequence and distinguished by lightweight identity-aware positional encodings. However, we identify a critical obstacle to this approach: the \textit{Reference Shortcut}. During training under standard noise schedules, the model can identify the matching reference by acoustic similarity to the noisy target, bypassing the text prompt entirely. We address this with a high-noise-biased timestep distribution that forces the model to rely on the text prompt for speaker assignment. We evaluate ScenA on the CoVoMix2-Dialogue benchmark, showing that it outperforms existing multi-speaker systems on speaker-binding metrics while generating rich conversational audio with overlapping speech, emotional vocalizations, and ambient sound. Our results demonstrate the advantage of using a general-purpose audio model conditioned on a free-form scene description, rather than passing structured dialog scripts through a speech-only pipeline.

In today's media-driven world, the exponential growth of video content across domains such as surveillance, education, and entertainment has made retrieving semantically relevant videos via natural language queries increasingly critical. Early video retrieval systems relied on handcrafted features or shallow cross-modal mappings, limiting their ability to capture complex semantics and temporal dynamics. While large-scale vision-language models have improved cross-modal alignment, challenges remain in modeling fine-grained temporal dependencies and nuanced linguistic structures. In this paper, we introduce DREAM: Dual-path Representation Enhancement and Alignment Model, a novel multimodal framework that addresses these limitations through enhanced visual and textual encoding. DREAM incorporates a hybrid language modeling strategy that combines masked and permuted language modeling objectives to capture both local and global linguistic semantics. On the visual side, we design a hierarchical vision encoder with cascaded group attention, which integrates spatial and temporal information through multi-stage token interaction and coarse-to-fine attention refinement. We validate DREAM through comprehensive evaluations on the widely-used MSRVTT, MSVD and LSMDC benchmark datasets, where it achieves new state-of-the-art R1 scores of 49.4%, 49.7% and 27.3%, respectively. Qualitative analyses further show the model's ability to maintain coherent attention across frames and align complex queries with dynamic video content. These findings underscore the effectiveness of hierarchical attention and dual-objective textual modeling in enabling robust, context-aware video retrieval, and pave the way for future research in advancing cross-modal representation learning.

Current automated pipelines for audio-visual Question Answering (QA) generally adopt a ``video-caption-QA'' paradigm. However, these methods typically segment videos into short clips and generate separate descriptions for audio and visual modalities. This decoupled processing severs inherent associations between sounds and their visual sources, while independent clip processing often causes inconsistent descriptions of the same entity across segments. Furthermore, coupling long-text comprehension and QA synthesis into a single step often restricts models to localized events, yielding questions lacking long-term temporal connections and deep cross-modal reasoning. To address these issues, we propose an automated data engine featuring two mechanisms: (1) \textbf{Entity-Anchored Video Scripting} transforms videos into structured scripts, comprising summaries, main entity lists, and segment-wise audio-visual descriptions. The entity list serves as a global prior to ensure cross-segment referential consistency and reconstruct audio-visual associations. (2) \textbf{Clue-Guided QA Generation} prompts models to first mine cross-segment, multimodal clues from the script, and subsequently generate QA pairs based on these high-value clues. Leveraging this pipeline, we construct the instruction-tuning dataset \textbf{OmniVideo-100K} and a human-verified test set, \textbf{OmniVideo-Test}. Fine-tuning VITA-1.5, Qwen2.5-Omni-7B and Qwen3-Omni-30B on OmniVideo-100K yields performance gains of up to 20.59% on OmniVideo-Test, demonstrating strong generalization (up to 12.64% improvements) across established benchmarks like Daily-Omni and JointAVBench.

Passive models for long video understanding typically rely on a "watch-it-all" paradigm, processing frames uniformly regardless of query difficulty, causing computational cost to grow with video duration. Although interactive frameworks have emerged, they often rely on global pre-scanning, and their context cost still scales with video length. We propose OmniAgent, the first native omni-modal agent that formulates video understanding as a POMDP-based iterative Observation-Thought-Action cycle. OmniAgent executes on-demand actions to selectively distill audio-visual cues into a persistent textual memory, effectively decoupling reasoning complexity from raw video duration. To operationalize this, we introduce (1) Agentic Supervised Fine-Tuning to bootstrap native active perception via best-of-N trajectory synthesis with dual-stage quality control, and (2) Agentic Reinforcement Learning with TAURA (Turn-aware Adaptive Uncertainty Rescaled Advantage), which leverages turn-level entropy to steer credit assignment toward pivotal discovery turns. Crucially, OmniAgent exhibits positive test-time scaling, where performance improves as the number of reasoning turns increases, validating the efficacy of active perception. Empirical results across ten benchmarks (e.g., VideoMME, LVBench) demonstrate that OmniAgent achieves state-of-the-art performance among open-source models. Notably, on LVBench, our 7B agent outperforms the 10$\times$ larger Qwen2.5-VL-72B (50.5% vs. 47.3%).

Unified multimodal models (UMMs) interleave generated ''visual thoughts'' (VTs) with text reasoning to improve spatial tasks. This incurs roughly an order-of-magnitude inference cost from multi-step diffusion. We find this cost yields limited direct benefit. On ThinkMorph, removing or noising VTs barely changes accuracy across nine benchmarks. Once rendered, attention concentrates on the VT regardless of content. Yet a KL diagnostic shows that conditioning on a privileged VT trace shifts the model's completion distribution. This suggests the generation pathway encodes useful reasoning beyond the rendered pixels. Motivated by this gap, we propose Visual On-Policy Self-Distillation(Visual-OPSD). Teacher and student share identical weights but differ in context: the teacher sees privileged VTs while the student sees only the question. Token-level JSD distillation on on-policy student trajectories transfers the teacher's reasoning to a text-only student. Across nine benchmarks, Visual-OPSD improves over its generative teacher by $+3.40$pp with $14.3\times$ speedup (10.0s vs. 142.8s per sample) and outperforms same-scale VLMs by $+63.83$pp on VSP. A Gaussian-noise control ($+0.40$pp vs. $+10.28$pp for real VTs) and $58.4\%$ closure of the KL gap confirm that gains come from the semantic content of the generation pathway.

Multimodal Information Extraction (MIE)-covering tasks such as Multimodal Named Entity Recognition (MNER), Relation Extraction (MRE), and Event Extraction (MEE)-is essential for understanding multimedia content but remains constrained by severe data scarcity. Although data augmentation is a promising remedy, existing approaches are impeded by coarse cross-modal alignment and fragmented, task-specific designs that fail to exploit shared semantic knowledge. To overcome these limitations, we introduce Semantic Anchor-aligned Multimodal Augmentation (SAMA), a unified framework for generating high-fidelity, task-aware synthetic data. SAMA constructs structured semantic anchors from ground-truth labels to guide a Collaborative Multi-Experts Multimodal Large Language Model (CME-MLLM), which integrates a Universal Adapter for shared semantics with Task-Specific Adapters to produce diverse yet constraint-compliant textual samples. For image synthesis, SAMA employs an Anchor-Preserving Diffusion mechanism that uses anchor-weighted prompts and latent conditioning to maintain critical semantic anchors while diversifying visual contexts. To eliminate the need for manual verification, SAMA further introduces a Dual-Constraint Filtering module that selects synthetic samples based on both cross-modal consistency and anchor fidelity. Extensive experiments across benchmark datasets for MNER, MRE, and MEE demonstrate that SAMA consistently outperforms state-of-the-art augmentation baselines under both fully supervised and low-resource settings, underscoring its versatility, robustness, and effectiveness.

Physical events are not understood by their names alone, but by the causal state changes that compose them. A clip-level label such as "bounce" can be correct while hiding the process that makes the event physically valid, from support loss and contact onset to rebound and settling. To make this hidden process explicit, we introduce Atomic Physical Transitions (APTs): minimal, temporally localized state changes that bind a visible cue to an active physical mechanism and before/after dynamical regimes. An APT chain represents a video as an ordered causal transition sequence rather than a single aggregate event label: event labels tell what happened; APT chains explain why it happened. To make APTs learnable by VLMs, we construct mixed-source APT data from human annotations and simulator ground truth, covering 14 transition types across contact, gravity, friction, and rotation/stability, with 27,303 timed instances over 1,246 trials. Using this data, we find that current VLMs miss transition-level physics, with zero-shot recall at most 14% and errors dominated by missed transitions. Direct fine-tuning on APT chains improves transition detection but causes event-level forgetting, indicating that the model learns a specialized answer format rather than a reusable physical representation. We therefore propose APT-Tune, a parameter-efficient recipe that teaches VLMs to use causal transitions without forgetting how to answer video questions. It combines image-pad-aware supervision, format-conditional co-training, and mechanism-conditioned domain-to-type decoding to make APT learning format-robust and physically grounded. With only 11 M LoRA parameters on Qwen3-VL-2B, APT-Tune substantially improves APT recall while also improving event-level video transfer. These results show that APTs are not a new answer format, but a human-aligned causal supervision signal for physical video understanding.

Traditional image captioning methods often struggle to generate comprehensive, context-rich descriptions, especially for details not directly observable from visual cues. To overcome this, we propose a novel retrieval-augmented image captioning framework that generates captions with deeper insights, such as object attributes, event context, and underlying significance, by leveraging external knowledge. Our approach features a hierarchical multi-modal article retrieval mechanism that moves beyond monolithic text entities. This retrieval considers article structure-aware features, including weighted textual components (e.g., headlines, body sections) and visual placement patterns, alongside multi-faceted similarity computations (content--visual, visual--visual, and discourse positioning). A subsequent contextual relevance refinement stage further enhances the retrieved information. The retrieved articles then serve as the knowledge base for caption generation: first, a VLM generates a concise image description; second, we segment relevant information from the retrieved articles based on this description; and finally, an LLM utilizes both the description and extracted knowledge to generate a comprehensive, contextually detailed caption. We participated in the ACM Multimedia EVENTA 2025 Challenge and achieved 5th place with an overall score of 0.2824 on the private test set of the OpenEvent-V1 dataset. Source code is publicly released at https://github.com/mf0212/EVENTA-Challange.

Domain adaptation remains a central challenge in 3D vision, especially for multimodal foundation models that align 3D point clouds with visual and textual data. While these models demonstrate strong general capabilities, adapting them to downstream domains with limited data often leads to overfitting and catastrophic forgetting. To address this, we introduce ReFine3D, a regularized fine-tuning framework designed for domain-generalizable tuning of 3D large multimodal models (LMMs). ReFine3D combines selective layer tuning with two targeted regularization strategies: multi-view consistency across augmented point clouds and text diversity through synonym-based prompts generated by large language models. Additionally, we incorporate point-rendered vision supervision and a test-time augmentation mechanism with confidence-based aggregation to further enhance robustness. Extensive experiments across different 3D domain generalization benchmarks show that ReFine3D improves base-to-novel class generalization by 1.36%, cross-dataset transfer by 2.43%, robustness to corruption by 1.80%, and few-shot accuracy by up to 3.11%, outperforming prior state-of-the-art methods with minimal added computational overhead.

Large Vision and Language Models (LVLMs) have advanced rapidly, yet European Portuguese (pt-PT) remains systematically underserved by existing open-source multimodal models, which either conflate it with Brazilian Portuguese or severely under-represent it in their training data mixes. We introduce AMALIA-VL, the first open-source instruction-tuned LVLM built natively for pt-PT, pairing a high-resolution vision encoder with dynamic image tiling and a fully open pt-PT-optimized language model via a learned connector. We contribute with a purposefully designed three-stage training process - vision-language alignment, general visual instruction tuning, and preference optimization - together with a pt-PT-centric multimodal data mix combining curated and translated public datasets with novel datasets that address the near-total absence of European Portuguese multimodal resources. Our evaluation shows that AMALIA-VL establishes a strong baseline for open-source pt-PT LVLMs.We will release model weights, training data, and construction pipelines along with machine-translated pt-PT evaluation benchmarks to help democratize pt-PT LVLM development.

Composed image retrieval (CIR) uses a reference image and a text modification to search for a target image. However, such queries often describe several possible images rather than one exact target, making the user's intent ambiguous. Recent methods address this by using conformal prediction to estimate ambiguity and by asking users clarifying text questions. However, these methods have two limitations: their coverage guarantee only holds at the first interaction, and text questions are often insufficient for resolving fine-grained visual differences such as appearance, attributes, or viewpoint. We propose CLARA, a clarification framework that resolves ambiguity by showing users a small panel of visual alternatives. Instead of answering text questions, the user simply selects the prototype image closest to the intended target. This provides a direct visual signal and avoids relying on a model to predict the user's answer. To maintain valid conformal guarantees across multiple interaction rounds, CLARA reweights calibration using the likelihood ratio induced by the user's selection. The displayed prototypes are also constrained to represent the current candidate set and are snapped to real corpus images, ensuring that generated images cannot artificially improve coverage. Experiments on open-domain and fashion benchmarks show that CLARA matches single-turn state-of-the-art retrieval performance, maintains nominal coverage across interaction rounds, and finds the intended target in fewer rounds than strong text-question baselines. Its advantage is especially clear when ambiguity involves viewpoint or fine-grained attributes, where visual clarification is more effective than textual questioning.

Image retrieval in crowded scenes is particularly challenging due to the salience bias of conventional visual encoders, which tend to focus on dominant objects while neglecting low-attention regions that are often crucial for fine-grained retrieval. We propose LARE (Low-Attention Region Encoding), a framework that explicitly models these overlooked regions. LARE adopts a dual-encoding strategy that encodes low-attention regions of an image and the full image in parallel, leading to more diverse and informative image embeddings. To evaluate image retrieval performance in challenging crowded scenes, we introduce Dense-Set, a challenging subset derived from COCO and Flickr30K. In this subset, images are re-captioned to provide richer descriptions of low-attention or previously overlooked regions. This dataset highlights the limitations of existing retrieval models and enables a more rigorous evaluation under densely crowded scene conditions. Experimental results demonstrate that the proposed framework improves retrieval performance by preserving subtle, non-dominant visual cues within the shared latent space.

Motion forecasting is central to visual intelligence: agents must anticipate how objects will move in order to plan actions, reason about physical interactions, and synthesize realistic futures. We argue that 3D points in world coordinates provide a general representation that is class-agnostic, view-stable, compact, and directly useful for downstream tasks. We formalize the task of goal-conditioned 3D point motion forecasting: given a short visual history, a set of 3D query points on an object of interest, and a language description of the intended goal, the model predicts the future 3D trajectory of each point. We introduce a full stack to study this task at scale: (1) MolmoMotion-1M is a large corpus of action-described, object-grounded 3D point trajectories annotated from 1.16M unconstrained videos; (2) PointMotionBench is a human-verified benchmark spanning 111 object categories and 61 motion types; and (3) MolmoMotion is a general motion forecasting model that supports both autoregressive coordinate prediction and flow-matching-based trajectory generation. MolmoMotion accurately predicts diverse motion patterns with different language instructions, and significantly outperforms existing motion prediction baselines on PointMotionBench. Finally, we show that the learned 3D motion prior transfers well to downstream applications: it improves training efficiency and generalization for robot manipulation, and its predicted trajectories provide effective motion guidance for generative models to synthesize videos with more realistic object motion.

Reinforcement learning has improved the reasoning ability of large language models, but applying outcome-only rewards to video multimodal large language models (Video-MLLMs) provides limited guidance on which visual evidence should support the answer. Inspired by multisensory integration, where consistent cues can enhance the salience and reliability of perceptual estimates, we introduce Consensus Frame GRPO (CF-GRPO), a temporal-annotation-free process-level reward framework for evidence-aware video reasoning. CF-GRPO constructs a consensus frame prior from intrinsic video cues, including temporal coverage, scene-transition cues, and query-conditioned visual relevance. It then computes a model-side frame-use score from visual and response representations and optimizes their agreement through the Consensus Frame Reward (CFR). With salience-aware sparse aggregation and distribution sharpening, CFR provides a high-contrast reward signal without requiring human temporal annotations. Experiments show that VideoCFR achieves competitive performance across complex video reasoning benchmarks and improves several metrics over representative Video-MLLM and RL baselines, while the consensus prior provides an interpretable view of the evidence frames emphasized during training. The implementation is available at https://github.com/1Pansy/VideoCFR.

Deploying multimodal foundation models as closed-loop policies increasingly requires conditioning actions on observations that are no longer visible. However, existing benchmarks either expose the full state, conflate hidden-state reconstruction with other agent skills, or test recall only after an episode has ended. We introduce RNG-Bench (Reconstructive Non-Markov Games), a benchmark suite designed to isolate a base model's ability to reconstruct past observations and act on them during multi-step interaction. RNG-Bench includes two complementary games: Matching Pairs, where card identities briefly revealed at specific locations must later be recalled, and 3D Maze, where egocentric views must be integrated into a spatial map. Both games are evaluated under a unified harness with three controlled difficulty axes: grid size, visual pattern, and observation modality. The benchmark further introduces a head-to-head duel protocol to control for instance-level variance and a Memory Gap metric that disentangles forgetting from poor action selection. The hardest configurations require contexts of roughly 128K tokens and 350 image inputs per episode, and remain far from saturated by frontier MLLMs. Memory Gap analysis shows that most residual errors stem from forgetting earlier observations rather than from suboptimal decision making. Finally, fine-tuning Qwen3.5-9B on optimal-policy rollouts and filtered model demonstrations improves performance on RNG-Bench and transfers to existing benchmarks without degrading general multimodal capability.

Embodied Vision-Language-Action (VLA) models are typically obtained by fine-tuning powerful pretrained VLMs on robotics data, yet it is unclear how much commonsense and factual knowledge they retain after adaptation. Failures on knowledge-sensitive tasks are ambiguous, conflating missing knowledge with poor generalization of low-level control. We introduce Act2Answer, a lightweight protocol that adapts VLM knowledge benchmarks to VLA evaluation by requiring agents to answer through action. Each question becomes a short tabletop episode where the agent performs a single object-placement action to select among candidate answers, yielding an action-grounded success rate with reduced control confounds. We curate a test suite of such environments across diverse commonsense and world-knowledge categories and introduce layerwise intent probing to localize answer-relevant information across the VLM backbone and action head. In a large-scale study of 7 VLA models and 9 VLM baselines, we systematically rank models across categories, finding that VLAs show solid performance on simple concepts while exhibiting larger gaps on richer semantic categories relative to their source VLMs, that VQA co-training is associated with better knowledge retention, and that answer-relevant signals peak in middle VLA layers but attenuate in upper layers. Act2Answer is available at https://tttonyalpha.github.io/act2answer/.

Establishing a universal benchmark for tactile representation learning in robotic manipulation remains challenging due to the diversity of tactile sensor designs, data formats, and robot embodiments. Rather than seeking to establish such, we explore a scalable and promising direction for future development: egocentric vision paired with full-hand tactile data. To this end, we introduce \textbf{HT-Bench}, a large-scale multi-task benchmark for dexterous full-hand tactile sensing, comprising 10M RGB frames and 7.8M tactile frames collected across 226 tasks. HT-Bench evaluates tactile representations from three key perspectives: whether they encode meaningful contact geometry, whether they can align tactile observations with visual information, and whether they generalize to unseen tasks. To assess these capabilities, HT-Bench includes four tasks: fine-grained tactile similarity retrieval, masked tactile inpainting, vision-to-tactile synthesis, and multimodal tactile frame prediction. We further propose \textbf{HandTouch}, a vector-quantized vision--tactile encoder that learns tactile representations through progressive spatial, cross-modal, and temporal training. Across HT-Bench, HandTouch consistently outperforms representative tactile encoder baselines, improving Recall@5 on fine-grained tactile similarity retrieval from 74.65\% to 85.23\%, reducing RMSE on masked tactile inpainting from 0.022 to 0.010, and increasing OOD cIoU on vision-to-tactile synthesis from 0.628 to 0.705. These results demonstrate the effectiveness of HandTouch and suggest that large-scale egocentric full-hand tactile data provides a scalable basis for evaluating and advancing tactile representation learning in dexterous manipulation.

Vision-Language Models (VLMs) enable robots to follow open-language instructions. However, dense VLM embeddings have shown to be noisy and lack spatial consistency. This is problematic for robotic applications, which require simultaneous reasoning over semantics and 3D space. We examine spatial structure across recent VLMs and propose ReSiReg, a feature reconstruction method that uses spatially consistent VLM intermediates to improve dense language-grounded retrieval. ReSiReg clusters intermediates into visual prototypes, derives their language descriptors, and reconstructs each patch as a soft mixture of prototype-level language embeddings. We evaluate quantitatively on OVSS and 3D mapping across backbones, and qualitatively in real-world manipulation scenes. Quantitative results show improved dense retrieval; manipulation scenes show more spatially consistent target activations. We further provide a compact 25M dense VLM for robotic applications, substantially smaller than and competitive with ViT-B baselines. Available at https://resireg.github.io

Training generalist Vision-Language-Action(VLA) models typically requires massive, diverse robotic datasets with high-fidelity action annotations. While egocentric human manipulation videos are abundant and capture significant environmental diversity, the absence of action labels makes them difficult to use in conventional training paradigms. To address this, we propose a latent-action-based framework designed to extract general action priors from unlabeled human videos. The architecture features a Hybrid Disentangled VQ-VAE that decouples motion dynamics from environmental backgrounds through physical masks, enabling the construction of a cross-embodiment action codebook. By pre-training on human videos with the codebook, the VLM backbone learns deep representations of action intent. For adaptation to specific embodiments, we introduce an intent-perception decoupling strategy where the VLM predicts the action intent while a separate frozen visual encoder provides state-specific features to the action expert, thereby reducing action hallucinations. Results in simulation and real-world environments show that our method, pre-trained exclusively on unlabeled human videos, performs competitively with state-of-the-art VLA models trained on massive annotated datasets, requiring only 50 trajectories for downstream adaptation.

Vision-language models (VLMs) are rapidly advancing toward sophisticated grounded structured visual reasoning. Training models for such advanced capabilities demands a new genre of data that seamlessly unifies spatial coordinates, open-vocabulary descriptions, structured attributes, and topological relationships into a singular representation. However, existing data annotation tools fundamentally fail to meet these intricate demands, suffering from three systematic bottlenecks: limited expressiveness, severe annotation-training decoupling, and poor data reusability. To bridge this infrastructure gap, we introduce an open-source annotation tool, ScreenAnnotator. First, we define a unified annotation atom schema that binds spatial, semantic, and structural primitives into a single unit. Second, we implement an on-policy annotation loop embedded with a Bayesian Annotation Verifier (BAV). Finally, we design a template-driven multi-task data synthesis process dynamically transforms static atoms into diverse multi-dimensional reasoning tasks, eliminating redundant re-annotation. The on-policy loop drives the annotation accept rate to nearly 100% on flowcharts and 77% on GUI screenshots, while steadily reducing per-image annotation time as labeled data accumulate. In the flowchart scenario, fine-tuning a VLM yields 76.1% average accuracy, which is a 35.1% point absolute gain. Our code is available at: https://github.com/WnQinm/Annotator.

While highlight detection for long-form videos is of great practical importance, most existing methods remain limited to short-form content, largely due to the absence of a suitable benchmark. To bridge this gap, we introduce SVHighlights, to the best of our knowledge, the first benchmark for highlight detection in extremely long sports videos, each exceeding one hour in duration, across multiple sports categories. SVHighlights is constructed from pairs of full-length sports videos and their corresponding official highlight videos using a dataset generation pipeline, enabling scalable label generation without conventional per-clip saliency annotation. The benchmark comprises 320 videos with an average duration of 2.00 hours and a total of 640.18 hours, substantially exceeding previous datasets. Existing methods also face fundamental challenges on long videos: models trained on short clips fail to generalize to hour-long content, and their clip-level scoring lacks the broader context needed to identify highlights. To address this and provide a strong baseline, we present TF-SELECTOR, a training-free segment-based approach that divides each video into context-aware segments by merging adjacent shots sharing the same semantic content, and predicts segment-level saliency scores using a large language model with multimodal inputs including visual captions, transcripts, and audio volume. Experiments demonstrate that TF-SELECTOR achieves superior performance across most metrics compared to Video Temporal Grounding (VTG)-tuned baselines, with improvements of +2.50 in HIT@1, +4.04 in HIT@K, and +2.95 in IoU. These results establish SVHighlights as a challenging testbed for long-form highlight detection and demonstrate that a simple segment-based strategy can effectively scale to hour-long videos.

On-policy self-distillation (OPSD) trains a model on its own rollouts and uses a frozen copy to provide dense token-level targets conditioned on a reference target. This works well for LLM reasoning, but a direct extension to multimodal large language models (MLLMs) can create a shortcut: the privileged target may guide tokens mainly based on the text reference target rather than the image. We propose ViGOS, a visually grounded OPSD framework for MLLM post-training. The student first writes a visual description and then reasons toward the final answer. For valid rollouts, an image-only perception teacher supervises the description, while a privileged reasoning teacher supervises the reasoning and final answer on the same student prefix. A reference teacher is used only for invalid rollouts to recover the output format. Across general vision-language, expert reasoning, visual math, spatial grounding, and visual-language-prior benchmarks, ViGOS keeps the main benefits of OPSD and improves image-grounded behavior in shortcut-prone settings.

Vision-language models (VLMs) are now widely used in downstream tasks. However, real-world applications often expose VLMs to distribution shifts induced by semantic variation (e.g., shape, size, and style). Robustness certification determines if a model's prediction changes when transformations are applied to its input. While most certification frameworks study geometric or pixel-level transformations over inputs, this work proposes a novel framework that enables certifying VLM robustness under semantic-level transformations. Leveraging the open-vocabulary capability of VLMs, we use text prompts as semantic proxies to construct transformations parameterized by an extent that controls the degree of semantic variation. By characterizing the VLM decision boundary in closed form, our framework quantitatively certifies extent intervals for which the predicted class remains unchanged under the semantic transformation. Our framework is the first to certify VLM robustness under semantic-level variations without requiring additional data for each variation, making it practical to apply. Experiments on both synthetic and real-world data show that our framework enables certifying robustness under diverse semantic variations across scenarios.

This paper presents an invertible neural network adapter for general robotic manipulation, designed to generate precise high-dimensional actions conditioned on multimodal observations, including visual, linguistic, and proprioceptive inputs, through a one-step denoising process. Built upon a flow-matching formulation, the proposed adapter effectively constrains the action generation trajectory within an invertible latent space, thereby enabling efficient and high-quality dexterous action synthesis with only a single inference step. Compared with conventional iterative flow-matching policies, the proposed framework substantially reduces inference complexity while maintaining strong action prediction accuracy and stability. Extensive experiments are conducted across a diverse set of simulation benchmarks and real-world robotic platforms to evaluate the effectiveness of the proposed method. Across simulation benchmarks, the proposed adapter consistently demonstrates superior or near state-of-the-art performance on a wide range of manipulation tasks. Furthermore, real-world experiments reveal a significant improvement in inference efficiency for vision-language-action (VLA) models, reducing the average inference latency from 110 ms to 61 ms while maintaining strong task performance.