Masked Motion Predictors are Strong 3D Action Representation Learners.
Yunyao MaoJiajun DengWengang ZhouYao FangWanli OuyangHouqiang LiPublished in: ICCV (2023)
Keyphrases
- recognizing human actions
- recognition of human actions
- action recognition
- view invariant
- motion estimation
- e learning
- learning environment
- image sequences
- human actions
- learning systems
- human motion
- motion field
- motion analysis
- collaborative learning
- image representation
- learning algorithm
- optical flow
- learning activities
- motion primitives
- learning community
- motion detection
- learning outcomes
- space time
- language learning
- learning resources
- camera motion
- learning scenarios
- motion parameters
- spatio temporal
- learning experience
- motion tracking
- configuration space
- motion model