Frame-Level Event Representation Learning for Semantic-Level Generation and Editing of Avatar Motion.
Ayaka IdenoTakuhiro KanekoTatsuya HaradaPublished in: ICMI (2023)
Keyphrases
- semantic level
- learning process
- expert systems
- learning algorithm
- context dependent
- perceptual information
- reinforcement learning
- motion parameters
- motion estimation
- human motion
- supervised learning
- image frames
- event detection
- video frames
- camera motion
- optical flow
- prior knowledge
- spatio temporal
- similarity measure
- image sequences