arXiv cs.AI· Dazhao Du, Shiyan Du, Jian Liu, Yongjian Yu, Bohai Gu, Tao Han, Jie Zhang, Hualuo Liu, Eric Liu, Yujia Zhang, Xi Chen, Song Guo·· 1 天前AI 评分30
CamChoreo：面向视觉语言模型的时间定位组合式相机运动理解基准
CamChoreo: Benchmarking Temporally Grounded Compositional Camera Motion Understanding for Vision Language Models
AI 导读
研究者提出 CamChoreo，首个要求模型定位运动一致区间并识别全部同时运动的相机运动理解基准，含 4,229 段单镜头视频与 20 个方向感知标签，近半片段为复合运动。
正文 · 原文
Authors:Dazhao Du, Shiyan Du, Jian Liu, Yongjian Yu, Bohai Gu, Tao Han, Jie Zhang, Hualuo Liu, Eric Liu, Yujia Zhang, Xi Chen, Song Guo
Abstract:Understanding camera motion is fundamental to video perception, with applications in spatial intelligence and controllable video generation. Existing camera motion evaluations typically assign labels to entire clips, overlooking within-shot changes and simultaneous movements. We therefore formulate camera motion understanding as temporally grounded, compositional recognition, requiring models to localize motion-consistent intervals and identify all active movements. We introduce CamChoreo, the first benchmark for this task, comprising 4,229 diverse single-shot clips with expert temporal annotations using 20 direction-aware labels. Nearly half of its segments contain compound camera motion. We also construct a separate training corpus of approximately 43K clips with the same annotation format. On CamChoreo, supervised fine-tuning (SFT) on this corpus achieves 62.1/64.1 micro F1 with 4B/8B backbones, outperforming advanced closed-source vision language models (VLMs). To complement the primarily semantic representations of VLM visual encoders with cross-frame geometry, we introduce CamInject, which injects features from a frozen 3D foundation model into the VLM. CamInject improves over SFT but requires running the costly 3D model at inference. We instead propose CamDistill, which distills the same geometric knowledge into lightweight camera tokens during training and removes the 3D model at inference. CamDistill improves micro F1 over same-data SFT by 5.4/3.7 points on 4B/8B backbones, closely approaching CamInject. CamDistill-8B also demonstrates strong zero-shot transfer on the external camera motion understanding benchmarks CameraBench and CameraMotionVQA. Together, these resources and CamDistill enable benchmarking and improving temporally grounded, compositional camera motion understanding.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2608.10932 [cs.CV] |
| (or arXiv:2608.10932v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2608.10932 arXiv-issued DOI via DataCite |
Submission history
From: Dazhao Du [view email]
[v1]
Tue, 11 Aug 2026 14:00:02 UTC (16,290 KB)
[v2]
Thu, 12 Nov 2026 07:40:23 UTC (16,733 KB)
来源：arXiv cs.AI · arxiv.org