Unifying Human Motion Synthesis and Style Transfer with Denoising Diffusion Probabilistic Models.
Ziyi ChangEdmund J. C. FindlayHaozheng ZhangHubert P. H. ShumPublished in: CoRR (2022)
Keyphrases
- denoising
- probabilistic model
- motion synthesis
- nonlinear diffusion
- human motion
- human movements
- image denoising
- graphical models
- natural images
- image processing
- motion capture data
- anisotropic diffusion
- image sequences
- expectation maximization
- diffusion process
- transfer learning
- latent variables
- action recognition
- image segmentation