Caption
-
Cross-Modal Latent Influence: Patch-Level Training Sample Attribution for Vision-Language Models
2026-10-09 18:04:04Laura Yao, Shixing Yu, Jie Peng, Promit Ghosal, Kyra Gan, Tianlong ChenAbstract
Explaining a vision--language model (VLM) prediction requires tracing training-example influence to the multimodal evidence that mediates it, yet existing attribution methods largely operate only at the sample level. Recent latent-mediated influence methods provide feature-level attribution in language models, but lack the multimodal and spatial structure needed to ground influence in VLMs. We introduce MOSAIC (Multimodal Orthogonalized Sparse Autoencoders for Influence Characterization). It learns a shared sparse image--text latent space whose concepts are 1) sparse within each modality, 2) supported on contiguous image regions, 3) aligned to the phrases that describe them, and 4) approximately decorrelated. It then applies EK-FAC to decompose training-example influence through these concepts and localize their contributions to specific text phrases and image regions. We evaluate MOSAIC on Qwen and Gemma across captioning, chest X-ray understanding, and visual question answering. The learned features are more localized and coherent without degrading task behavior, and ablating influence-ranked latents changes model outputs more than ablating latents ranked by activation or frequency. We find that visual evidence can substantially reorganize training-data attribution while preserving more of the shared latent structure. MOSAIC thus connects where model behavior comes from with what multimodal evidence carries that influence.
Abstract (translated)
URL
https://arxiv.org/abs/2610.14190
PDF
-
Images Talk Back: Guiding MMDiTs by Suppressing Image-to-Text Feedback
2026-10-09 17:58:15Yucheng Yang, Xiaoxiao Sun, Shusheng Yang, Jiakang Chen, Serena Yeung-LevyAbstract
Multimodal diffusion transformers (MMDiTs) allow the evolving image to update internal text representations through joint attention. These representations then guide subsequent image updates, raising a question: does this feedback help correct deviations from the prompt, or preserve them? We find that image-to-text (I2T) feedback can preserve the objects and attributes already represented in the evolving image, including errors. Transplanting I2T messages can rescue failures or disrupt successes. At matched intervention strengths, suppressing the influence of I2T feedback improves generation, whereas opposite controls degrade it. These findings motivate Semantic Feedback Guidance (SFG), which suppresses the influence of I2T feedback by guiding away from an auxiliary branch that strengthens this feedback and weakens text-to-image conditioning. The in-place variant, iSFG, directly suppresses I2T feedback and strengthens text-to-image conditioning, requiring no extra model evaluations. Temporal interventions motivate running the SFG branch during the first 20% of image sampling steps, requiring about $1.2\times$ the base NFE without classifier-free guidance (CFG). Experiments across three image-generation backbones show gains in alignment and preference-model scores. On FLUX, adding SFG to CFG raises GenEval from $0.6661$ to $0.7479$; on SD3.5M, iSFG adds $+1.33$ HPSv2.1 over CFG at zero extra NFE. The control also transfers to text-to-video and image-to-video generation, improving VBench scores. Code is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13941
PDF
-
Hide and Seek: HIDE Concepts to SEEK explanations for black box models
2026-10-09 17:56:59Bhanu Tokas, Rahul Nair, Rockwell Jackson, Hannah KernerAbstract
Interpreting machine learning model predictions is a central challenge, particularly for black-box vision models. Concept-based attribution methods interpret models using human-understandable concepts. Existing concept-based methods can only interpret classification and regression models. They require access to model weights or large concept-annotated datasets, and most methods cannot provide image-level attribution. We propose Hide and Seek (HnS), a new method for concept-level attribution that does not require access to model weights or labeled concept datasets. HnS can interpret a wide range of vision models. Given a concept and image set, HnS retrieves concept-aligned images using a vision-language model, localizes the concept within each image, masks the localized region, and measures attribution from the change in the black-box model's output between the original and masked images. This enables faithful attribution by directly measuring the model's behavior and supporting local attributions. We validated HnS on CelebA and COCO-Stuff. Compared to existing methods, HnS showed better agreement with ground-truth attributions. We demonstrated the applicability of HnS on diverse tasks, including image captioning and satellite image segmentation, where no existing concept attribution method can be used. We provided our anonymous code in the supplementary material.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13935
PDF
-
Decoding Register Tokens Preserves Global Scene Details
2026-10-09 09:46:53David Jacob Drexlin, Julius Hense, Jonas DippelAbstract
Pretrained vision transformers learn structured image representations spanning spatial patch tokens and non-spatial tokens. Recent analyses suggest that non-spatial register tokens encode global scene properties such as illumination, blur, and style. Yet patch-based image decoders discard these outputs, leaving the decoder to recover image appearance from patches alone. We investigate whether retaining register features improves decoding when the complete spatial patch representation is already available. We study this question using representation autoencoders (RAEs) with frozen register-equipped encoders. Our decoder processes four register tokens together with all patch tokens through joint self-attention. Appearance probes, paired reconstruction-error analysis, and fixed-decoder register substitutions examine which global image properties are accessible from registers and how their values affect reconstructed pixels. For generation, we train a multi-head attention pooling (MHAP) predictor to recover register features from sampled patches, keeping the pretrained diffusion model and its samples unchanged. On ImageNet-1K, our ViT-B decoders achieve mean reconstruction FIDs of 0.472 with DINOv2 and 0.442 with DINOv3, reductions of 9.5% and 5.2% relative to their respective patch-only baselines. The corresponding PSNRs are 19.73 and 20.54 dB. For generation, our DINOv2 models achieve a mean FID of 1.641, 9.8% below the patch-only baseline on identical sampled patches.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13299
PDF
-
Vision Transformers Learn Word Meaning from Images Alone
2026-10-09 09:00:08Anders Gj{\o}lbye, Sebastian Ray Mason, Stella Frank, Lars Kai HansenAbstract
Can a vision model learn what words mean from pixels alone? We show that the two largest DINOv3 vision transformers organize words rendered in images by word meaning, despite being pretrained using self-supervision on images only, without paired captions, class labels, or other language supervision. We derive word representations from embedding differences between word images and matched controls, then test whether semantically related words are neighbors within each model's embedding space, while controlling for spelling similarity. We find that brand names in the same industry group together, as do nouns from the same semantic domain. Country-capital pairs exhibit consistent vector offsets, similarly to word embeddings trained on text. These word representations also align with the representations of the visual object categories they name. DINOv3-H and DINOv3-7B are, to our knowledge, the first self-supervised vision models to show substantial semantic structure for rendered words. Comparisons with text and speech encoders reveal shared structure among representations of the same words. Contrastive alignment with either encoder further strengthened DINOv3-H's word-level semantic structure, while leaving the other finetuned vision models largely unchanged. Image-only training can therefore teach vision models not just what words look like, but also how their meanings relate.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13252
PDF
-
BenCoDiC: A Large-Scale Contextual Cyberbullying Corpus for Bengali Political News Discourse FULL VERSION
2026-10-09 07:16:29Md. Nesarul Hoque, Rudra Pratap Deb Nath, Umme Salma, Shima Chakraborty, Md. Hanif SeddiquiAbstract
Cyberbullying is a growing concern in online communities, yet existing Bengali corpora remain limited in contextual information, annotation transparency, scale, accessibility, and support for multidimensional analysis, thereby hindering robust detection and comprehensive study of harasser behaviors and networks. To address these gaps, we introduce the Bengali Contextual and Diverse Cyberbullying (\textit{BenCoDiC}) corpus, a publicly available contextual cyberbullying resource for Bengali political news discourse comprising 50,250 textual instances collected from 100 YouTube videos and their associated discussions. The corpus preserves caption--comment--reply relationships and is organized through a nine-dimensional star schema covering users, videos, textual content, context, time, engagement, political parties, target identities, and language categories. Five AI chatbots were initially evaluated for annotation but did not meet the predefined reliability criterion. Consequently, a two-tier human annotation process with expert verification was adopted, achieving Fleiss' $\kappa=0.86$ and Krippendorff's $\alpha=0.86$. Statistical analyses characterize the corpus, while transformer-based experiments show that incorporating conversational context yields modest but consistent improvements over isolated-text classification. \textit{BenCoDiC} supports context-aware cyberbullying detection and perpetrator-oriented behavioral and network analysis. The corpus is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13121
PDF
-
WaveMark: Robust Wavelet-Domain Image Watermarking for Diffusion Models
2026-10-09 03:46:45Jie Cao, Qi Li, Chaklam Cheong, Yutong Jin, Enze Yu, Zelin Zhang, Jianbing NiAbstract
Image watermarking enables provenance tracing for photorealistic images generated by diffusion models. However, state-of-the-art (SOTA) removal attacks can suppress embedded signals with little perceptual degradation. This vulnerability limits the effectiveness of watermarks for intellectual property protection and misuse attribution. We propose WaveMark, a novel wavelet-domain watermarking method that embeds messages in a diffusion model's initial noise. To support robust recovery from perturbed carriers, WaveMark distributes keyed replicas of message bits across spatial positions, wavelet orientations, and scales. Each carrier groups coefficients from four latent channels and encodes a randomized bit through magnitude projection, providing an explicit decoding margin. We evaluate WaveMark with SD2.1 on SDP and MS-COCO Captions under 14 attacks. WaveMark achieves 100.0% bit accuracy (BA) without attacks, while maintaining image quality comparable to the evaluated baselines. Under SOTA MarkNull and MarkNull-A, it retains over 88% BA, outperforming all baselines.
Abstract (translated)
URL
https://arxiv.org/abs/2610.12952
PDF
-
A Lightweight Velocity Correction Framework for Improving Instruction Following in Few-Step Text-to-Image Generation
2026-10-09 02:43:49Yating Gao, Zhenlong Zhu, Wen Zihao, Lenan Zhang, Jiexiong Wang, Yang Yang, Guanjun Jiang, Yixiong Zou, Takuya MaekawaAbstract
Few-step text-to-image generators produce high-quality images with few network forward passes, yet struggle with complex prompts involving multiple subjects, interactions, spatial relations, counting, and attribute binding. Can these errors be reduced without changing the generator's parameters or increasing its forward-pass count? We propose a lightweight correction framework for a frozen 6B-parameter generator. The generator predicts velocities that determine how its intermediate image representation is updated; a corrector with fewer than 20M parameters learns residual adjustments at selected sampling steps. Because earlier corrections change the states encountered at later steps, our On-Trajectory Flow-Bridge Supervision derives residual targets at the states actually reached during corrected generation. A single corrector is shared across selected steps and runs in parallel with the generator. Only the corrector is trained, requiring neither reward optimization nor re-distillation. To our knowledge, this is the first framework for lightweight velocity-residual adaptation of fully frozen few-step text-to-image generators that preserves their original sampling schedule and forward-pass count. On our target complex-scene distribution, the framework achieves a +17.60-percentage-point NetWin over frozen Z-Image-Turbo in pairwise instruction-following evaluation, with only 26.9 ms (0.84%) additional end-to-end latency per image and no increase in backbone evaluations.
Abstract (translated)
URL
https://arxiv.org/abs/2610.12899
PDF
-
Unsupervised Multimodal Reward Modeling
2026-10-08 20:38:25Yueying Li, Jingxuan Fan, Kiant\'e Brantley, Sham M. Kakade, Hanlin ZhangAbstract
Vision-language reward models are trained on annotated preference pairs, which are expensive to collect, can disagree, and are ranked by benchmark accuracy. We build one with no human preference label: a linear head on an intermediate layer of a frozen Qwen3-VL backbone, trained to score a web caption above a perturbed version of the same caption; labeled data enters only to select the layer. The layer is the decisive choice, worth 5.8 points over the last layer, and one recipe transferred without re-tuning beats or at least matches GPT-4o on three benchmarks at 32B, 6.6 points ahead on VL-RewardBench, from a few thousand parameters at about a thousandth of fine-tuning compute. At an equal pair budget, perturbations teach a frozen head 3 to 5 points more than curated RLAIF-V pairs. Because the backbone is frozen, the objective is convex, and the perturbation sampler specifies the reward, which the benchmark does not identify: two heads trained on disjoint halves of the same perturbations tie on the benchmark and differ by 6.4 points in hallucination judging. Retraining the head on the policy's own verifier-labeled candidates cuts caption hallucination by 5.1 CHAIR points and lifts image-generation pass rate by 5.4, while scoring lower on every reward benchmark. Together, an intermediate-layer probe, free same-image perturbations, and a verifier-driven bootstrap turn give a recipe for multimodal reward modeling that needs no annotator, costs a fraction of fine-tuning, and is specified by the comparisons it is trained on rather than by a benchmark rank.
Abstract (translated)
URL
https://arxiv.org/abs/2610.12707
PDF
-
OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning
2026-10-08 17:59:21Zhongyu Yang, Jiale Tao, Ruitao Chen, Zuhao Yang, Yingfang Yuan, Xueliang Zhao, Auden, Kai Wang, Shuai Shao, Biao Wang, Steve Yves, Qinglin LuAbstract
Multimodal large language models (MLLMs) are rapidly evolving toward continuous audio--visual reasoning, creating an urgent need for evaluations that expose their capability limits. Audio--visual captioning is an ideal diagnostic task, yet current benchmarks face a coupled trade-off: whole-caption scores provide coverage without localization, local probes provide localization without coverage, and unconstrained LLM judges introduce instability. We introduce OmniCapBench (Omni-Video Caption Benchmark), a benchmark that reframes audio--visual caption evaluation as a deep-structured diagnostic framework. OmniCapBench shifts the prediction target from free-form text to sets of atomic, verifiable evaluation units across three tracks: entity references, visual shots, and audio events, enabling reliable scoring with deterministic constraint checks and localized LLM-based semantic comparisons. With 786 densely annotated videos, OmniCapBench effectively distinguishes MLLM perception errors, including temporal grounding failures, identity drift, cross-modal misalignment, and hallucinated descriptions. Evaluating frontier MLLMs reveals strong local perception but weak long-horizon audio--visual reasoning, particularly in identity drift and cross-modal misalignment, providing a fine-grained roadmap for omnimodal development.
Abstract (translated)
URL
https://arxiv.org/abs/2610.12458
PDF
-
FastBench: Can Streaming VLMs Perceive High-Dynamic Real-World Streams?
2026-10-08 17:55:20Yuxuan Hu, Weikang Shi, Yang Bo, Xudong Lu, Xintong Guo, Shuhan Li, Yuyang He, Huankang Guan, Peiwen Sun, Yunqiao Yang, Wenbo Li, Rui Liu, Hongsheng LiAbstract
Streaming Video Large Language Models (VLMs) enable continuous video understanding, yet existing benchmarks focus on low-dynamic scenarios. Under bounded context budgets, models must balance temporal history, spatial resolution, and temporal granularity; sparse sampling at 1--2 FPS misses fast events. We introduce FastBench to evaluate high-dynamic perception in real-world video streams. Its trajectory-grounded pipeline combines QA generation from high-FPS clips, filtering of questions answerable at 2 FPS, answer verification using SAM3 and CoTracker3 trajectories, and three rounds of human inspection. FastBench contains 306 QA pairs across eight domains, six capabilities, and forward, instant, and backward temporal scopes, with human-annotated evidence intervals. We also present ProactiveFrame, a training-free baseline that adjusts incoming frame rates through text tokens. A dual-tier sliding window retains recent high-FPS observations while downsampling older ones into sparse history. Experiments reveal substantial limitations: the strongest model, Gemini-3.5-Flash, scores only 50.7%. Denser sampling improves Qwen3-VL-8B from 32.9% at 2 FPS to 44.6% at 24 FPS, but gains saturate as history is compressed. ProactiveFrame outperforms sparse uniform sampling by 5.4 and 1.5 percentage points, yet remains well below oracle-guided focusing, showing that current VLMs struggle to determine from the stream alone when finer temporal perception is needed. FastBench provides a testbed for high-dynamic streaming video understanding. Code and data: this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.12427
PDF
-
DataVista: Diagnosing Multimodal LLMs on Data Video Understanding
2026-10-08 14:02:35Yupeng Xie, Zhenyang Wang, Jiayi Zhu, Yinghao Tang, Zhouan Shen, Yiyu Chen, Yuyu LuoAbstract
Data video is a media form that integrates data visualization with video narrative, widely adopted in news reporting and business analysis. Compared with general video understanding, data video understanding places greater emphasis on accurately reading data from animated charts, integrating evidence across charts and time, and understanding how narrative organization and visual design communicate information. Yet existing benchmarks target either general videos or static charts, and data video understanding has not been systematically evaluated. We present DataVista, the first benchmark for data video understanding, containing 961 real-world data videos and 6,775 evaluation questions organized under a three-level progressive capability framework (data perception, temporal reasoning, narrative understanding) with 10 fine-grained question types across five topic domains. Systematic evaluation of 19 mainstream MLLMs shows that the best-performing model, Gemini-3.1-Pro, achieves 70.0% overall accuracy, still far below human expert performance, with models performing worst on Causal Reasoning and Narrative Structure. Increasing frame counts and adding subtitles mainly benefit data perception and temporal reasoning, with limited gains in narrative understanding. Further analysis of model responses identifies typical failure modes in chart reading, evidence judgment, and instruction understanding. The benchmark is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11993
PDF
-
From Suppression to Repair: Mitigating Object Hallucination in Large Vision-Language Models via Localized Distribution Alignment
2026-10-08 12:25:50Chen Zhao, Xingping Dong, Jiachun Shi, Liang Peng, Chong Wang, Zhen Lei, Ran He, Bo DuAbstract
Object hallucination remains a major obstacle for large vision-language models (LVLMs) to generate reliable content. An intuitive mitigation strategy is to suppress hallucination-related components in hidden representations. However, these components may also contain useful information, and suppressing them can weaken the model's multimodal capabilities. In this paper, we propose ResOT, a training-free method that repairs representations at inference time through localized distribution alignment. Specifically, ResOT projects dominant hallucinated directions away from the faithful subspace, forming a low-dimensional residual subspace for intervention. Within this subspace, ResOT uses Gaussian optimal transport (OT) to align the hallucinated distribution with the faithful one. The resulting map defines repair targets with minimal changes to the original representations. At inference, ResOT adaptively controls how far each token state moves toward its OT target. Experiments on three representative LVLMs show that ResOT substantially reduces object hallucination while improving image caption quality and multimodal performance across multiple benchmarks. Code will be released.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11826
PDF
-
Towards Unified Evaluation of Prompt Enhancers for Video Generation
2026-10-08 11:31:13Yawen Shao, Yubo Zhu, Ziyun Dai, Zixun Fang, Kai Zhu, Zeyinzi Jiang, Yufeng Ai, Siyang Sun, Haolan Xue, Yu Shang, Yuxiang Bao, Zoubin Bi, Jingming Luo, Jie Xiao, Chaojie Mao, Zhehan Kan, Hongchen Luo, Yu Liu, Sheng Zhong, Wei Tong, Xueyang Fu, Yang Cao, Wei Zhai, Zheng-Jun ZhaAbstract
Modern video generators can realize increasingly complex visual narratives, positioning the prompt enhancer (PE) as a critical bridge from concise user instructions and multimodal references to structured cinematic plans. However, existing PE evaluation relies on rendered videos, imposing substantial computational and human costs, slowing PE training and iteration, and conflating PE quality with downstream generator behavior. To address this gap, we introduce PEBench, the first unified benchmark for direct PE evaluation across text-to-video, image-to-video, and reference-to-video prompt enhancement. It comprises 1,100 expert-verified cases and 1,005 visual assets, spanning 35 fine-grained tasks with diverse temporal, cinematic, audiovisual, and multi-reference requirements. In addition, we develop PEBench evaluation, an evidence-grounded framework that combines modality-aware fact extraction with rubric-based assessment across 24 criteria. Our systematic evaluation of representative open- and closed-source PE methods reveals an emerging shift from fine-grained descriptive expansion toward intent-preserving cinematic planning, while the caption-reconstruction and forward-refinement methods show complementary strengths in cinematic coverage and semantic fidelity or internal coherence, respectively. Human validation shows that PEBench scores align closely with expert judgments of enhanced prompts and downstream videos from Wan3.0 and MiniMax-H3, indicating that prompt-level evaluation reliably reflects downstream utility.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11736
PDF
-
SV-TAD: Native Sparse Convs for Efficient Temporal Action Detection
2026-10-08 09:31:38Ricardo Pizarro, Roberto Valle, Jos\'e M. Buenaposada, Luis M. Bergasa, Luis BaumelaAbstract
To adapt billion-parameter Vision Transformers for long-video understanding, recent methods freeze the backbone and train lightweight convolutional modules. While effective for parameter-efficient training, existing adapters do not reduce inference-time computation, leaving scalability with respect to video length largely unaddressed. Token selection can reduce attention cost by pruning redundant tokens, but it breaks the spatial grid structure required by convolutional adapters. This forces an expensive dense reconstruction, nullifying much of the potential speedup. We address this by introducing native sparse 2D convolutions, a primitive that allows these adapters, for the first time, to operate directly and efficiently on dynamically pruned token sets. We integrate this primitive into SV-TAD, an adapter framework for temporal action detection, reducing VideoMAEv2-L computation by up to 64% and achieving 2.2x faster inference, while maintaining state-of-the-art accuracy on THUMOS-14 and ActivityNet-1.3. When scaled to InternVideoNext-L, our approach surpasses the previous state of the art at roughly half its computational cost. Moreover, the sparse formulation naturally supports auxiliary task tokens, which improves fine-grained assembly detection on ATTACH.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11579
PDF
-
Beyond Speech Captions: Speech-Rewarded Style Planning for Conversational Text-to-Speech
2026-10-08 08:12:28Shiao Zhu, Lianbo Liu, Sizhen Lyu, Yuzhe Wang, Sheng Li, Takahiro ShinozakiAbstract
Natural-language style descriptions provide an interpretable interface between large language models (LLMs) and controllable text-to-speech (TTS). However, using descriptions as pseudo-labels compresses target acoustics into text, and descriptive fidelity need not imply effective control of a particular synthesizer. We empirically show that speech-text alignment only weakly predicts downstream acoustic similarity among candidate instructions for the same utterance. We therefore propose Speech-Rewarded Style Planning (SRSP), which trains a text-based style planner through a frozen downstream TTS model. Given dialogue history and response text, the planner generates candidate instructions and is optimized with group-relative policy optimization (GRPO), using the teacher-forced likelihood of target speech tokens as the reward. On an English subset of the ISCSLP 2026 CoT-TTS corpus, SRSP achieves higher speech-style and emotion similarity to target speech and lower mel-cepstral distortion than the Base LLM and target-audio-informed captioning baselines. LLM-based expressive speech evaluation further shows gains over all baselines in contextual appropriateness and reference consistency.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11461
PDF
-
ProtoSemImage: Image-Valued Prototypes with Deformable Row Alignment for Interpretable Document Classification
2026-10-08 08:12:22Mohammad Zare, Pirooz ShamsinejadbabakiAbstract
Prototypes in classification models are almost always vectors, and a vector has no readable form. This paper asks what happens when a prototype is an image. Documents give the question a natural form, because a document can be rendered as a multi-channel image in which every token becomes a pixel, so a class representative can take the same shape and the same channel semantics as the inputs it stands for. ProtoSemImage represents each class by one or more visual archetypes: prototype images in a four-channel HSV space whose channels carry named linguistic factors. A Skip-Gram objective learns that color space end to end through a four-dimensional bottleneck, discourse boundary rows become differentiable typed difference rows, and classification reduces to 2D visual template matching: a deformable row alignment between a document image and the archetype bank, in the spirit of dynamic time warping. Because the match is a spatial pattern comparison rather than a linear readout, the model reports where an input departs from its archetype and along which channel, and a generative head decodes each archetype back into text. The image representation works: it beats an otherwise identical model with vector prototypes in all three paired seeds, by between 4.3 and 11.8 points on a ten-class task. The distance-based matching does not. A diagnostic that keeps the representation fixed and swaps only the classifier recovers the sequence baselines, which locates a 20.6-point shortfall in the matching rather than in the color compression, and a benchmark built so that a pair of documents shares a bag of words and differs only in arrangement confirms the layout-preservation it was designed for. We report both directions, because for a representation whose whole purpose is inspect ability, the failure modes are as informative as the gains.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11460
PDF
-
SepGen: Multi-Stem Audio-Video Separation and Generation in a Single Model
2026-10-08 06:51:59Aviad Dahan, Rajaei Khatib, Yonatan Bitton, Idan Szpektor, Lior Wolf, Raja GiryesAbstract
A 4D audio-visual scene comprises a video, the dynamic geometry it depicts, and the sound sources that populate it, each with its own position and trajectory. Rendering such a scene from a novel viewpoint requires that every source be available as an individual waveform, so that it can be localized in the scene and propagated to the observer before the signals are mixed. Joint audio-video generators can synthesize the video and its soundtrack, but the soundtrack is emitted as a single audio-mix in which the sources are not individually accessible. We present SepGen, which extends a pretrained audio-video generator to emit the video, the mixed soundtrack, and one waveform per captioned source in one joint sampling run. SepGen supports two complementary modes: generation and separation. In generation mode, each source caption specifies what its stem contains. A two-speaker dialogue, for example, comes out as one stem per speaker in the original turn order. In separation mode, the input audio-mix remains clean while the captions specify what to extract, so the model can decompose a recording from a free-text description. We evaluate generation on scenes synthesized from text, and separation on scenes rendered by other generators and on real recordings of speech, music, and sound effects. Given an audio-mix and captions that carry the spoken lines, SepGen outperforms language-conditioned separators, most clearly on speech, and it keeps the lead when the lines are removed from the captions. Code, checkpoints, and datasets are available at this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.11361
PDF
-
GATOR: Generative and Agentic 3D Object Reconstruction From Casual Images
2026-10-08 04:13:04Qirui Wu, Stan Birchfield, Hesam Rabeti, Angel X. Chang, Bowen WenAbstract
Reconstructing complete, scene-aligned 3D objects from casual images requires integrating sparse, uncertain observations and inferring surfaces hidden by occlusions. We present GATOR, a generative and agentic framework that recovers textured object assets and their scene-relative pose from one or more images. Our local modality mixer couples patch-aligned RGB, target-mask, and pointmap features before cross-view reasoning, preserving scene context while distinguishing the target from its surroundings. Text-guided semantic conditioning complements these spatial cues with category names and object captions through stage-specific adapters for structure, geometry, and appearance generation. The generated asset initializes a multimodal agent, providing instance-specific geometry and pose for targeted structural and texture refinement through an observation-guided edit-render-review loop. Across synthetic objects, cluttered tabletops, and indoor scenes, GATOR achieves strong geometric and appearance fidelity while recovering scene-relative pose from sparse observations. Time-budget comparisons and scene-level simulation further demonstrate the reconstruction efficiency and simulation readiness. Project page: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.11215
PDF
-
VAMR: Multi-Question Agentic Reasoning for Efficient Long-Form Video Understanding
2026-10-08 03:26:54Runquan Gui, Hanzhu Chen, Zehao Wang, Hanxin Zhu, Xin Li, Zhibo ChenAbstract
Long-form video understanding often involves multiple questions about different aspects of the same recording. Yet existing video agents typically process each question through an isolated tool-use trajectory. This repeatedly restarts video exploration and memory construction, missing opportunities to acquire evidence jointly and progressively build a shared understanding that supports the complete question set. We introduce \textbf{VAMR} (\textbf{V}ideo \textbf{A}gent for \textbf{M}ulti-Question \textbf{R}easoning), which coordinates all questions about a video through one shared tool-use trajectory. At each round, a persistent policy model can invoke tools for one or more unresolved questions and submit answers for questions with sufficient evidence. Question-conditioned visual perception retrieves fine-grained clues for several questions in one call, while layered multi-question memory integrates reusable context into a shared video story and preserves separate evidence for individual questions. After supervised fine-tuning initializes this interaction protocol, we propose question-horizon policy optimization (\qhpo) to optimize shared trajectories in which questions progress and finish at different rounds. Specifically, a question-level critic estimates the value of each active question, while round alignment maps each question advantage to the rounds that directly serve it before the aligned advantages are aggregated to optimize the shared actor. Across LVBench, Video-Holmes, and LongVideoBench, VAMR achieves the highest accuracy overall and the fewest reasoning rounds among iterative methods. On LVBench, it reaches 62.1\% accuracy, exceeding VideoARM by \textbf{4.3} points while reducing reasoning rounds and processed frames by \textbf{85.9\%} and \textbf{61.4\%}.
Abstract (translated)
URL
https://arxiv.org/abs/2610.11171
PDF
