Running on Zero
Agents
SkeletonDiffusion
💻
Generate future 3D motion videos from an input clip
None defined yet.
This is the official Huggingface space of the CVPR2025 paper Nonisotropic Gaussian Diffusion for Realistic 3D Human Motion Prediction.
Code is available on GitHub. Paper on arXiv: arxiv.org/abs/2501.06035.
Interactive Demo from video and pretrained models are available in this space.