Transformers have emerged as a universal backbone across 3D perception, video generation, and world models for autonomous driving and embodied AI, where understanding camera geometry is essential for grounding visual observations in three-dimensional space. However, existing camera encoding methods often rely on simplified pinhole assumptions, restricting generalization across the diverse intrinsics and lens distortions in real-world cameras. We introduce **Relative Ray Encoding**, a geometry-consistent representation that unifies complete camera information, including 6-DoF poses, intrinsics, and lens distortions. To evaluate its capability under diverse controllability demands, we adopt camera-controlled text-to-video generation as a testbed task. Within this setting, we further identify pitch and roll as two components effective for **Absolute Orientation Encoding**, enabling full control over the initial camera orientation. Together, these designs form **UCPE (Unified Camera Positional Encoding)**, which integrates into a pretrained video Diffusion Transformer through a lightweight spatial attention adapter, adding **less than 1% trainable parameters** while achieving state-of-the-art camera controllability and visual fidelity. To facilitate systematic training and evaluation, we construct a large video dataset covering a wide range of camera motions and lens types. Extensive experiments validate the effectiveness of UCPE in camera-controllable video generation and highlight its potential as a general camera representation for Transformers across future multi-view, video, and 3D tasks.
论文检索
输入标题、作者或关键词，从 2,006 篇学术成果中精准定位
ReMoRa: Multimodal Large Language Model based on Refined Motion Representation for Long-Video Understanding
PDF ↗While multimodal large language models (MLLMs) have shown remarkable success across a wide range of tasks, long-form video understanding remains a significant challenge.In this study, we focus on video understanding by MLLMs.This task is challenging because processing a full stream of RGB frames is computationally intractable and highly redundant, as self-attention have quadratic complexity with sequence length.In this paper, we propose ReMoRa, a video MLLM that processes videos by operating directly on their compressed representations.A sparse set of RGB keyframes is retained for appearance, while temporal dynamics are encoded as a motion representation, removing the need for sequential RGB frames.These motion representations act as a compact proxy for optical flow, capturing temporal dynamics without full frame decoding.To refine the noise and low fidelity of block-based motions, we introduce a module to denoise and generate a fine-grained motion representation.Furthermore, our model compresses these features in a way that scales linearly with sequence length.We demonstrate the effectiveness of ReMoRa through extensive experiments across a comprehensive suite of long-video understanding benchmarks.ReMoRa outperformed baseline methods on multiple challenging benchmarks, including LongVideoBench, NExT-QA, and MLVU. Our project page is available at https://remora-v1rcm.kinsta.page/.
Video understanding with multimodal large language models (MLLMs) remains challenging due to the long token sequences of videos, which contain extensive temporal dependencies and redundant frames.Existing approaches typically treat MLLMs as passive recognizers, processing entire videos or uniformly sampled frames without adaptive reasoning.Recent agent-based methods introduce external tools, yet still depend on manually designed workflows and perception-first strategies, resulting in inefficiency on long videos. We present EVA, an Efficient Reinforcement Learning framework for End-to-End Video Agent, which enables planning-before-perception through iterative summary-plan-action-reflection reasoning.EVA autonomously decides what to watch, when to watch, and how to watch, achieving query-driven and efficient video understanding.To train such agents, we design a simple yet effective three-stage learning pipeline--comprising supervised fine-tuning (SFT), Kahneman-Tversky Optimization (KTO), and Group Relative Policy Optimization (GRPO)--that bridges supervised imitation and reinforcement learning.We further construct high-quality datasets for each stage, supporting stable and reproducible training.We evaluate EVA on six video understanding benchmarks, demonstrating its comprehensive capabilities. Compared with existing baselines, EVA achieves a substantial improvement of 6--12% over general MLLM baselines and a further 1--3% gain over prior adaptive agent methods.
This paper addresses the critical and underexplored challenge of long video understanding with low computational budgets.We propose LongVideo-R1, an active, reasoning-equipped multimodal large language model (MLLM) agent designed for efficient video context navigation, avoiding the redundancy of exhaustive search.At the core of LongVideo-R1 lies a reasoning module that leverages high-level visual cues to infer the most informative video clip for subsequent processing.During inference, the agent initiates traversal from top-level visual summaries and iteratively refines its focus, immediately halting the exploration process upon acquiring sufficient knowledge to answer the query.To facilitate training, we first extract hierarchical video captions from CGBench, a video corpus with grounding annotations, and guide GPT-5 to generate 33K high-quality chain-of-thought-with-tool trajectories. The LongVideo-R1 agent is fine-tuned upon the Qwen-3-8B model through a two-stage paradigm: supervised fine-tuning (SFT) followed by reinforcement learning (RL), where RL employs a specifically designed reward function to maximize selective and efficient clip navigation.Experiments on multiple long video benchmarks validate the effectiveness of name, which enjoys superior tradeoff between QA accuracy and efficiency. All curated data and source code are provided in the supplementary material and will be made publicly available.
Segmenting long-form videos into semantically coherent scenes is a fundamental task in large-scale video understanding. Existing encoder-based methods are limited by visual-centric biases, classify each shot in isolation without leveraging sequential dependencies, and lack both narrative understanding and explainability. In this paper, we present Scene-VLM, the first fine-tuned vision-language model (VLM) framework for video scene segmentation. Scene-VLM jointly processes visual and textual cues including frames, transcriptions, and optional metadata to enable multimodal reasoning across consecutive shots. The model generates predictions sequentially with causal dependencies among shots and introduces a context-focus window mechanism to ensure sufficient temporal context for each shot-level decision. In addition, we propose a scheme to extract confidence scores from the token-level logits of the VLM, enabling controllable precision-recall trade-offs that were previously limited to encoder-based methods. Furthermore, we demonstrate that our model can be aligned to generate coherent natural-language rationales for its boundary decisions through minimal targeted supervision. Our approach achieves state-of-the-art performance on standard scene segmentation benchmarks. On MovieNet, for example, Scene-VLM yields significant improvements of +6 AP and +13.7 F1 over the previous leading method.
Multi-view camera systems enable rich observations of complex real-world scenes, and understanding dynamic objects in multi-view settings has become central to various applications. Point tracking serves as a key mechanism for capturing dynamic motion. However, conventional single-view approaches often fail due to the limited geometric information available in monocular video, which becomes a critical bottleneck for multi-view extension. In this work, we present MV-TAP, a robust point tracker that tracks points across multi-view videos of dynamic scenes by leveraging cross-view information. MV-TAP utilizes camera geometry and a cross-view attention mechanism to aggregate spatio-temporal information across views, enabling more complete and reliable trajectory estimation in multi-view videos. To support this task, we construct a large-scale synthetic training dataset and real-world evaluation sets tailored for multi-view tracking. Extensive experiments demonstrate that MV-TAP outperforms existing point-tracking methods on challenging benchmarks, establishing an effective baseline for advancing research in multi-view point tracking.
Despite progress in video understanding, current MLLMs struggle with counting tasks. Existing benchmarks are limited by short videos, close-set queries, lack of clue annotations, and weak multimodal coverage. In this paper, we introduce CG-AV-Counting, a manually-annotated clue-grounded counting benchmark with 1,027 multimodal questions and 5,845 annotated clues over 497 long videos. It supports both black-box and white-box evaluation, serving as a comprehensive testbed for both end-to-end and reasoning-based counting. To explore ways to improve model's counting capability, we propose AV-Reasoner, a model trained with GRPO and curriculum learning to generalize counting ability from related tasks. AV-Reasoner achieves SOTA results across multiple benchmarks, demonstrating the effectiveness of reinforcement learning. However, experiments reveal that on out-of-domain benchmarks, reasoning in the language space offers limited performance gains, suggesting the need for more robust cross-domain reasoning mechanisms.
Current vision-language pre-training (VLP) paradigms excel at global scene understanding but struggle with instance-level reasoning due to global-only supervision. We introduce InstAP, an Instance-Aware Pre-training framework that jointly optimizes global vision-text alignment and fine-grained, instance-level contrastive alignment by grounding textual mentions to specific spatial-temporal regions. To support this, we present InstVL, a large-scale dataset (2 million images, 50,000 videos) with dual-granularity annotations: holistic scene captions and dense, grounded instance descriptions. On the InstVL benchmark, InstAP substantially outperforms existing VLP models on instance-level retrieval, and also surpasses a strong VLP baseline trained on the exact same data corpus, isolating the benefit of our instance-aware objective. Moreover, instance-centric pre-training improves global understanding: InstAP achieves competitive zero-shot performance on multiple video benchmarks, including MSR-VTT and DiDeMo. Qualitative visualizations further show that InstAP localizes textual mentions to the correct instances, while global-only models exhibit more diffuse, scene-level attention.
Physical AI aims to develop models that can perceive and predict real-world dynamics; yet, the extent to which current multi-modal large language models and video generative models support these abilities is insufficiently understood. We introduce Physical AI Bench (PAI-Bench), a unified and comprehensive benchmark that evaluates perception and prediction capabilities across video generation, conditional video generation, and video understanding, comprising 2,808 real-world cases with task-aligned metrics designed to capture physical plausibility and domain-specific reasoning. Our study provides a systematic assessment of recent models and shows that video generative models, despite strong visual fidelity, often struggle to maintain physically coherent dynamics, while multi-modal large language models exhibit limited performance in forecasting and causal interpretation. These observations suggest that current systems are still at an early stage in handling the perceptual and predictive demands of Physical AI. In summary, PAI-Bench establishes a realistic foundation for evaluating Physical AI and highlights key gaps that future systems must address.
SEASON: Mitigating Temporal Hallucination in Video Large Language Models via Self-Diagnostic Contrastive Decoding
PDF ↗Video Large Language Models (VideoLLMs) have shown remarkable progress in video understanding. However, these models still struggle to effectively perceive and exploit rich temporal information in videos when responding to user queries. Therefore, they often generate descriptions of events that are temporal inconsistent or causally implausible, causing severe hallucination issues. While most prior studies have focused on spatial hallucinations (e.g. object mismatches), temporal reasoning in video understanding remains relatively underexplored. To address this issue, we propose Self-Diagnostic Contrastive Decoding (SEASON), a training-free method that adaptively enhances temporal and spatial faithfulness for each output token. It achieves this by dynamically diagnosing each token's hallucination tendency and applying adaptive contrastive decoding against its corresponding temporal and spatial negatives. Extensive experiments demonstrate that SEASON outperforms all existing training-free hallucination mitigation approaches on three hallucination examination benchmarks, while further improves VideoLLMs across four general video understanding benchmarks. Our project page: https://chriswu018.github.io/season/
Test-time adaptation for video understanding, which enables vision-language models (VLMs) to generalize to downstream tasks such as action recognition, has demonstrated substantial value in real-world applications. Existing memory-based methods typically build a visual cache from high-confidence test videos and perform inference via a two-step modality mapping chain, i.e., vision-vision and vision-text. However, due to the asymmetry of the two mappings, the chain exhibits non-transitivity, hindering the generalization of VLMs. To this end, we propose a novel training-free Condensed Dynamic Adapter ConDA for action recognition, which leverages vision-text alignment to guide vision-vision alignment. It first selects semantic patches based on the semantic activation probability obtained from the vision-text alignment (Probability-based Semantic Patch Selection, PSPS), and then adaptively constructs spatial-temporal video tubes based on patch-level visual similarity (Adaptive Tube Construction, ATC). We conduct extensive experiments on seven benchmarks with different backbones and baselines. The quantitative results demonstrate that ConDA is compatible with arbitrary VLM and generalizes well across complex scenarios, such as long-term and egocentric scenarios. In addition, qualitative analyses showcase the interpretability of ConDA in terms of capturing semantic cues.
Discrete video VAEs underpin modern text-to-video generation and video understanding systems, yet existing tokenizers typically learn visual codebooks at a single scale with limited vocabularies and shallow language supervision, leading to poor cross-modal alignment and zero-shot transfer. We introduce PyraTok, a language-aligned pyramidal tokenizer that learns semantically structured discrete latents across multiple spatial-temporal resolutions. PyraTok builds on a pretrained video VAE and a novel Language aligned Pyramidal Quantization (LaPQ) module that discretizes encoder features at several depths using a shared large binary codebook, yielding compact yet expressive video token sequences. To tightly couple visual tokens with language, PyraTok jointly optimizes multi-scale text-guided quantization and a global autoregressive objective over the token hierarchy. Across ten benchmarks, PyraTok delivers state-of-the-art (SOTA) video reconstruction, consistently improves text-to-video quality, and sets new SOTA zero-shot performance on video instance segmentation, temporal action localization, and video understanding, scaling robustly to up to 4K/8K resolutions.
Dense Video Object Captioning (DVOC) is the task of jointly detecting, tracking, and captioning object trajectories in a video, requiring the ability to understand spatio-temporal details and describe them in natural language.Due to the complexity of the task and the high cost associated with manual annotation, previous approaches resort to training strategies with limited data, potentially leading to suboptimal performance.To circumvent this issue, we propose to generate captions about spatio-temporally localized entities leveraging a state-of-the-art VLM, and extend the LVIS and LV-VIS datasets with our synthetic captions (LVISCap and LV-VISCap). Moreover, we introduce an end-to-end model, CaptionFormer, capable of jointly detecting, segmenting, tracking and captioning object trajectories.With pretraining on LVISCap and LV-VISCap, CaptionFormer achieves state-of-the-art DVOC results on three existing benchmarks, VidSTG, VLN and BenSMOT.
Long video understanding presents significant challenges for vision-language models due to extremely long context windows. Existing solutions relying on naive chunking strategies with retrieval-augmented generation, typically suffer from information fragmentation and a loss of global coherence. We present HAVEN, a unified framework for long-video understanding that enables coherent and comprehensive reasoning by integrating audiovisual entity cohesion and hierarchical video indexing with agentic search. First, we preserve semantic consistency by integrating entity-level representations across visual and auditory streams, while organizing content into a structured hierarchy spanning global summary, scene, segment, and entity levels. Then we employ an agentic search mechanism to enable dynamic retrieval and reasoning across these layers, facilitating coherent narrative reconstruction and fine-grained entity tracking. Extensive experiments demonstrate that our method achieves good temporal coherence, entity consistency, and retrieval efficiency, establishing a new state-of-the-art with an overall accuracy of 84.1% on LVBench. Notably, it achieves outstanding performance in the challenging reasoning category, reaching 80.1%. These results highlight the effectiveness of structured, multimodal reasoning for comprehensive and context-consistent understanding of long-form videos.
Reliable 4D object detection, which refers to 3D object detection in streaming video, is crucial for perceiving and understanding the real world. Existing open-set 4D object detection methods typically make predictions on a frame-by-frame basis without modeling temporal consistency, or rely on complex multi-stage pipelines that are prone to error propagation across cascaded stages. Progress in this area has been hindered by the lack of large-scale datasets that capture continuous reliable 3D bounding box (b-box) annotations. To overcome these challenges, we first introduce DA4D, a large-scale 4D detection dataset containing over 280k sequences with high-quality b-box annotations collected under diverse conditions. Building on DA4D, we propose DetAny4D, an open-set end-to-end framework that predicts 3D b-boxes directly from sequential inputs. DetAny4D fuses multi-modal features from pre-trained foundational models and designs a geometry-aware spatiotemporal decoder to effectively capture both spatial and temporal dynamics. Furthermore, it adopts a multi-task learning architecture coupled with a dedicated training strategy to maintain global consistency across sequences of varying lengths. Extensive experiments show that DetAny4D achieves competitive detection accuracy and significantly improves temporal stability, effectively addressing long-standing issues of jitter and inconsistency in 4D object detection. Data and code will be released upon acceptance.
DocPrune: Efficient Document Question Answering via Background, Question, and Comprehension-aware Token Pruning
PDF ↗Recent advances in vision-language models have shown strong performance across diverse multimodal tasks, including document question answering that leverages structured visual cues from text, tables, and figures. However, unlike natural images, document images contain large backgrounds and only sparse supporting evidence, leading to the waste of substantial computational resources, especially for long documents. We observe that existing token reduction methods for natural images and videos fall short in utilizing the structural sparsity unique to documents. To address this, we propose DOCPRUNE, a training-free document token pruning framework designed for efficient long document understanding. The proposed method preserves only the essential tokens for the task while removing unnecessary ones, such as background or question-irrelevant tokens. Moreover, it automatically selects the appropriate layers to initiate token pruning based on the model's level of comprehension. Our experiments on the M3DocRAG benchmark show that DOCPRUNE improves throughput by 3.0x and 3.3x in the encoder and decoder, respectively, while boosting the F1 score by +1.0, achieving both higher accuracy and efficiency without any additional training.
Sound effects build an essential layer of multimodal storytelling, shaping the emotional atmosphere and the narrative semantics of videos. Despite recent advancement in video-text-to-audio (VT2A), the current formulation faces three key limitations: (1) an imbalance between visual and textual conditioning that leads to visual dominance; (2) the absence of a concrete definition for fine-grained controllable generation; (3) weak instruction understanding and following, as existing datasets rely on brief categorical tags. To address these limitations, we introduce EchoFoley (Event-Centric Hierarchical cOntrol Foley), a new task designed for video-grounded sound generation with both event-level local control and hierarchical semantic control. Our symbolic representation for sounding events specifies when, what, and how each sound is produced within a video or instruction, enabling fine-grained controls like sound generation, insertion, and editing. To support this task, we construct EchoFoley-6k, a large-scale, expert-curated benchmark containing over 6,000 video-instruction-annotation triplets and 42,000 fine-grained sounding event annotations.Building upon this foundation, we propose EchoVidia, a sounding-event-centric agentic generation framework with slow-fast thinking strategy. Experiments show that EchoVidia surpasses recent VT2A models by 40.7% in controllability and 12.5% in perceptual quality.
Understanding real-world videos such as movies requires integrating visual and dialogue cues. Yet existing VideoQA benchmarks struggle to capture this multimodal reasoning and, given the difficulty of evaluating free-form answers, largely resort to simple multiple choice questions. We introduce a novel open-ended multimodal VideoQA benchmark, MovieRecapsQA, created using movie recap videos -- a distinctive type of YouTube content that summarizes a film via a voiceover description of key clips from the movie (recap video). From the transcribed voiceover (recap summary) of 60 recap videos, we generate 8.2K questions along with the necessary "facts" expected in each answer; the former facilitates the creation of questions that require mutimodal reasoning and the latter allow the construction of a reference-free evaluation metric that can be applied to open-ended responses. To our knowledge, this is the first reference-free open-ended VideoQA benchmark. The benchmark allows each question to be evaluated in different input video settings: given (a) the full-length movie, (b) the full ( 11 min) recap video (visual only), (c) 14 min of aligned movie scenes, i.e, movie scenes relevant to the question, and (d) 1.2 min of aligned recap video scenes. In all cases, the text of any associated movie dialogue is provided. Each question is categorized by the modality required to answer it---visual, dialogue, or both---enabling fine-grained evaluation of multimodal capabilities. We benchmark (setting (d)) seven state-of-the-art MLLMs and find that (i) only our reference-free metric produces meaningful human-aligned model separation; (ii) vision-centric questions yield the lowest scores across all models; (iii) removing visual input often improves model factuality; and (iv) the primary bottleneck is visual perception, not visual reasoning.
Multimodal retrieval is the task of aggregating information from queries across heterogeneous modalities to retrieve desired targets. State-of-the-art multimodal retrieval models can understand complex queries, yet they are typically limited to two modalities: text and vision. This limitation impedes the development of universal retrieval systems capable of comprehending queries that combine more than two modalities. To advance toward this goal, we present OmniRet, the first retrieval model capable of handling complex, composed queries spanning three key modalities: text, vision, and audio. Our OmniRet model addresses two critical challenges for universal retrieval: computational efficiency and representation fidelity. First, feeding massive token sequences from modality-specific encoders to Large Language Models (LLMs) is computationally inefficient. We therefore introduce an attention-based resampling mechanism to generate compact, fixed-size representations from these sequences. This shared module is designed to maintain representational diversity and generalization capabilities while remaining sensitive to modality-specific information. Second, compressing rich omni-modal data into a single embedding vector inevitably causes information loss and discards fine-grained details. We propose Attention Sliced Wasserstein Pooling to preserve these fine-grained details, leading to improved omni-modal representations. OmniRet is trained on an aggregation of approximately 6 million query-target pairs spanning 30 datasets. We benchmark our model on 13 retrieval tasks and a MMEBv2 subset. Our model demonstrates significant improvements on composed query, audio and video retrieval tasks, while achieving on-par performance with state-of-the-art models on others. Furthermore, we curate a new Audio-Centric Multimodal Benchmark (ACM). This new benchmark introduces two critical, previously missing tasks--composed audio retrieval and audio-visual retrieval--to more comprehensively evaluate a model's omni-modal embedding capacity. We believe our benchmark will facilitate the development of universal retrieval systems.
We introduce the "single-life" learning paradigm, where we train a distinct vision model exclusively on egocentric videos captured by one individual. We leverage the multiple viewpoints naturally captured within a single life to learn a visual encoder in a self-supervised manner. Our experiments demonstrate three key findings. First, models trained independently on different lives develop a highly aligned geometric understanding. We demonstrate this by training visual encoders on distinct datasets each capturing a different life, both indoors and outdoors, as well as introducing a novel cross-attention-based metric to quantify the functional alignment of the internal representations developed by different models. Second, we show that single-life models learn generalizable geometric representations that effectively transfer to downstream tasks, such as depth estimation, in unseen environments. Third, we demonstrate that training on up to 30 hours from one week of the same person's life leads to comparable performance to training on 30 hours of diverse web data, highlighting the strength of single-life representation learning. Overall, our results establish that the shared structure of the world, both leads to consistency in models trained on individual lives, and provides a powerful signal for visual representation learning.