DiffSHEG: A Diffusion-Based Approach for Real-Time Speech-driven Holistic 3D Expression and Gesture Generation.
Junming ChenYunfei LiuJianan WangAiling ZengYu LiQifeng ChenPublished in: CoRR (2024)
Keyphrases
- real time
- facial animation
- continuous stream
- multimodal interfaces
- low cost
- hand movements
- data driven
- control system
- vision system
- high speed
- speech recognition
- audio visual
- sign language
- automatic speech recognition
- anisotropic diffusion
- speech signal
- generation process
- spoken language
- scale space
- hidden markov models