LGTM: Local-to-Global Text-Driven Human Motion Diffusion Model.
Haowen SunRuikun ZhengHaibin HuangChongyang MaHui HuangRuizhen HuPublished in: CoRR (2024)
Keyphrases
- human motion
- diffusion model
- motion capture
- human body
- optical motion capture
- image sequences
- information diffusion
- video sequences
- anisotropic diffusion
- body parts
- motion recognition
- spatio temporal
- text mining
- human movements
- diffusion process
- motion sequences
- human actions
- viewpoint
- motion estimation
- motion capture data
- computer vision
- visual data
- image registration
- image processing