Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.13326 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914329096355840 |
|---|---|
| author | Hu, Xirui Ding, Yanbo Wang, Jiahao Shi, Tingting Wang, Yali Zhi, Guo Zhi Zhang, Weizhan |
| author_facet | Hu, Xirui Ding, Yanbo Wang, Jiahao Shi, Tingting Wang, Yali Zhi, Guo Zhi Zhang, Weizhan |
| contents | Character image animation, which synthesizes videos of reference characters driven by pose sequences, has advanced rapidly but remains largely limited to single-human settings. Existing methods struggle to generalize to multi-humanoid scenarios, which involve diverse humanoid forms, complex interactions, and frequent occlusions. We address this gap with two key innovations. First, we introduce unified motion representations that extract identity-agnostic motions and explicitly bind them to corresponding characters, enabling generalization across diverse humanoid forms and seamless extension to multi-humanoid scenarios. Second, we propose a holistic 4D-anchored paradigm that constructs a shared 4D space to fuse motion representations with video latents, and further reinforces this process with hierarchical 4D-level supervision to better handle interactions and occlusions. We instantiate these ideas in MotionWeaver, an end-to-end framework for multi-humanoid image animation. To support this setting, we curate a 46-hour dataset of multi-human videos with rich interactions, and construct a 300-video benchmark featuring paired humanoid characters. Quantitative and qualitative experiments demonstrate that MotionWeaver not only achieves state-of-the-art results on our benchmark but also generalizes effectively across diverse humanoid forms, complex interactions, and challenging multi-humanoid scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_13326 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MotionWeaver: Holistic 4D-Anchored Framework for Multi-Humanoid Image Animation Hu, Xirui Ding, Yanbo Wang, Jiahao Shi, Tingting Wang, Yali Zhi, Guo Zhi Zhang, Weizhan Computer Vision and Pattern Recognition Character image animation, which synthesizes videos of reference characters driven by pose sequences, has advanced rapidly but remains largely limited to single-human settings. Existing methods struggle to generalize to multi-humanoid scenarios, which involve diverse humanoid forms, complex interactions, and frequent occlusions. We address this gap with two key innovations. First, we introduce unified motion representations that extract identity-agnostic motions and explicitly bind them to corresponding characters, enabling generalization across diverse humanoid forms and seamless extension to multi-humanoid scenarios. Second, we propose a holistic 4D-anchored paradigm that constructs a shared 4D space to fuse motion representations with video latents, and further reinforces this process with hierarchical 4D-level supervision to better handle interactions and occlusions. We instantiate these ideas in MotionWeaver, an end-to-end framework for multi-humanoid image animation. To support this setting, we curate a 46-hour dataset of multi-human videos with rich interactions, and construct a 300-video benchmark featuring paired humanoid characters. Quantitative and qualitative experiments demonstrate that MotionWeaver not only achieves state-of-the-art results on our benchmark but also generalizes effectively across diverse humanoid forms, complex interactions, and challenging multi-humanoid scenarios. |
| title | MotionWeaver: Holistic 4D-Anchored Framework for Multi-Humanoid Image Animation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.13326 |