LaserHuman: Language-guided Scene-aware Human Motion Generation in Free Environment
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914722233712640 |
|---|---|
| author | Cong, Peishan Wang, Ziyi Dou, Zhiyang Ren, Yiming Yin, Wei Cheng, Kai Sun, Yujing Long, Xiaoxiao Zhu, Xinge Ma, Yuexin |
| author_facet | Cong, Peishan Wang, Ziyi Dou, Zhiyang Ren, Yiming Yin, Wei Cheng, Kai Sun, Yujing Long, Xiaoxiao Zhu, Xinge Ma, Yuexin |
| contents | Language-guided scene-aware human motion generation has great significance for entertainment and robotics. In response to the limitations of existing datasets, we introduce LaserHuman, a pioneering dataset engineered to revolutionize Scene-Text-to-Motion research. LaserHuman stands out with its inclusion of genuine human motions within 3D environments, unbounded free-form natural language descriptions, a blend of indoor and outdoor scenarios, and dynamic, ever-changing scenes. Diverse modalities of capture data and rich annotations present great opportunities for the research of conditional motion generation, and can also facilitate the development of real-life applications. Moreover, to generate semantically consistent and physically plausible human motions, we propose a multi-conditional diffusion model, which is simple but effective, achieving state-of-the-art performance on existing datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_13307 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LaserHuman: Language-guided Scene-aware Human Motion Generation in Free Environment Cong, Peishan Wang, Ziyi Dou, Zhiyang Ren, Yiming Yin, Wei Cheng, Kai Sun, Yujing Long, Xiaoxiao Zhu, Xinge Ma, Yuexin Computer Vision and Pattern Recognition Language-guided scene-aware human motion generation has great significance for entertainment and robotics. In response to the limitations of existing datasets, we introduce LaserHuman, a pioneering dataset engineered to revolutionize Scene-Text-to-Motion research. LaserHuman stands out with its inclusion of genuine human motions within 3D environments, unbounded free-form natural language descriptions, a blend of indoor and outdoor scenarios, and dynamic, ever-changing scenes. Diverse modalities of capture data and rich annotations present great opportunities for the research of conditional motion generation, and can also facilitate the development of real-life applications. Moreover, to generate semantically consistent and physically plausible human motions, we propose a multi-conditional diffusion model, which is simple but effective, achieving state-of-the-art performance on existing datasets. |
| title | LaserHuman: Language-guided Scene-aware Human Motion Generation in Free Environment |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.13307 |