MotionAtlas: Detailed Region Captioning for Motion-Centric Videos

• URL: http://arxiv.org/abs/2606.29531v1
• Date: Sun, 28 Jun 2026 17:54:55 GMT
• Title: MotionAtlas: Detailed Region Captioning for Motion-Centric Videos
• Abstract summary: MotionAtlas is a system for captioning of motion-centric videos.<n>It generates precise descriptions of motion within the target region.<n> MotionAtlas-4B surpasses Qwen3-VL-4B by an average of 5.2 percentage points across general motion benchmarks.
• Score: 32.07943965658894
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We propose MotionAtlas, a system for detailed captioning of motion-centric videos, comprising (1) a dedicated human-annotated benchmark, (2) a scalable, high-quality pipeline to construct training samples, and (3) a family of powerful Video-MLLMs. Unlike conventional global motion captioning datasets, we focus on region-aware motion captioning: given a video and a spatiotemporal mask, the model generates precise descriptions of motion within the target region, thereby alleviating visual clutter and motion entanglement and enabling reliable, quantifiable evaluation. Concretely, we first build MotionAtlas-Bench, a comprehensive benchmark comprising 2,073 multiple-choice questions, meticulously annotated for a curated set of high-quality, motion-centric videos, to evaluate fine-grained motion understanding of the objects in question. Second, we design a rigorous and scalable data pipeline that leverages self-bootstrap refinement to suppress fine-grained hallucinations, yielding 159k high-quality motion captioning data. Third, we design a tailored training data composition strategy, which achieves consistent and substantial performance gains across diverse baseline Video-MLLMs, including Molmo2 and Qwen3-VL. For instance, MotionAtlas-4B surpasses Qwen3-VL-4B by an average of 5.2 percentage points across general motion benchmarks. The benchmark, dataset, and code have been released.

Related papers

• Feeling the Space: Egomotion-Aware Video Representation for Efficient and Accurate 3D Scene Understanding [9.774606705982874]
  This paper significantly enhances Large Language Models (MLLMs) with egomotion modality data.<n>By grounding visual content in physical egomotion trajectories, Motion-MLLM can reason about absolute scale and spatial relationships across the scene.<n>Compared to state-of-the-art (SOTA) methods based on video frames and explicit 3D data, Motion-MLLM exhibits similar or even higher accuracy with significantly less overhead.
  arXiv  Detail & Related papers  (2026-03-18T17:42:49Z)
• MeViS: A Multi-Modal Dataset for Referring Motion Expression Video Segmentation [126.77662882743168]
  We introduce MeViS, a dataset containing 33,072 human-annotated motion expressions in both text and audio.<n>We benchmark 15 existing methods across 4 tasks supported by MeViS.<n>We propose an approach LMPM++ for RVOS/AVOS/RMOT that achieves new state-of-the-art results.
  arXiv  Detail & Related papers  (2025-12-11T18:59:44Z)
• FoundationMotion: Auto-Labeling and Reasoning about Spatial Movement in Videos [109.99404241220039]
  We introduce FoundationMotion, a fully automated data curation pipeline that constructs large-scale motion datasets.<n>Our approach first detects and tracks objects in videos to extract their trajectories, then leverages these trajectories and video frames with Large Language Models.<n>We fine-tune open-source models including NVILA-Video-15B and Qwen2.5-7B, achieving substantial improvements in motion understanding without compromising performance.
  arXiv  Detail & Related papers  (2025-12-11T18:53:15Z)
• Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance [107.25252623824296]
  Wan-Move is a framework that brings motion control to video generative models.<n>Our core idea is to make the original condition features motion-aware for guiding video.<n>Through scaled training, Wan-Move generates 5-second, 480p videos whose motion controllability rivals Kling 1.5's commercial Motion Brush.
  arXiv  Detail & Related papers  (2025-12-09T16:13:55Z)
• DynamicEval: Rethinking Evaluation for Dynamic Text-to-Video Synthesis [17.750053029702222]
  Existing text-to-video (T2V) evaluation benchmarks, such as VBench and EvalCrafter, suffer from two limitations.<n>We introduce DynamicEval, a benchmark consisting of systematically curated prompts emphasizing dynamic camera motion.<n>For background scene consistency, we obtain the interpretable error maps based on the Vbench motion smoothness metric.<n>Our proposed metrics achieve stronger correlations with human preferences at both the video level and the model level.
  arXiv  Detail & Related papers  (2025-10-08T18:41:04Z)
• VimoRAG: Video-based Retrieval-augmented 3D Motion Generation for Motion Language Models [110.32291962407078]
  VimoRAG is a video-based retrieval-augmented motion generation framework for motion large language models.<n>We develop an effective motion-centered video retrieval model and mitigate the issue of error propagation caused by suboptimal retrieval results.<n> Experimental results show that VimoRAG significantly boosts the performance of motion LLMs constrained to text-only input.
  arXiv  Detail & Related papers  (2025-08-16T15:31:14Z)
• MOVE: Motion-Guided Few-Shot Video Object Segmentation [25.624419551994354]
  This work addresses motion-guided few-shot video object segmentation (FSVOS)<n>It aims to segment dynamic objects in videos based on a few annotated examples with the same motion patterns.<n>We introduce MOVE, a large-scale dataset specifically designed for motion-guided FSVOS.
  arXiv  Detail & Related papers  (2025-07-29T17:59:35Z)
• VideoMolmo: Spatio-Temporal Grounding Meets Pointing [66.19964563104385]
  VideoMolmo is a model tailored for fine-grained pointing of video sequences.<n>A novel temporal mask fusion employs SAM2 for bidirectional point propagation.<n>To evaluate the generalization of VideoMolmo, we introduce VPoMolS-temporal, a challenging out-of-distribution benchmark spanning five real-world scenarios.
  arXiv  Detail & Related papers  (2025-06-05T17:59:29Z)
• Learning to Segment Rigid Motions from Two Frames [72.14906744113125]
  We propose a modular network, motivated by a geometric analysis of what independent object motions can be recovered from an egomotion field. It takes two consecutive frames as input and predicts segmentation masks for the background and multiple rigidly moving objects, which are then parameterized by 3D rigid transformations. Our method achieves state-of-the-art performance for rigid motion segmentation on KITTI and Sintel.
  arXiv  Detail & Related papers  (2021-01-11T04:20:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.