Talking Head Generation with Probabilistic Audio-to-Visual Diffusion Priors.
Zhentao YuZixin YinDeyu ZhouDuomin WangFinn WongBaoyuan WangPublished in: ICCV (2023)
Keyphrases
- visual information
- learned from training data
- visual data
- cross modal
- visual features
- low level
- generative model
- bayesian framework
- probabilistic model
- audio visual
- multimedia
- prior probabilities
- visual perception
- prior knowledge
- bayesian networks
- uncertain data
- generation process
- high level
- real time
- anisotropic diffusion
- prior information
- maximum a posteriori
- video indexing and retrieval
- multimodal information