View-Invariant, Occlusion-Robust Probabilistic Embedding for Human Pose.
Ting LiuJennifer J. SunLong ZhaoJiaping ZhaoLiangzhe YuanYuxiao WangLiang-Chieh ChenFlorian SchroffHartwig AdamPublished in: CoRR (2020)
Keyphrases
- human pose
- view invariant
- action recognition
- human actions
- human motion
- body parts
- human detection
- human body
- pose estimation
- activity recognition
- motion capture data
- computer vision
- probabilistic model
- motion capture
- multi view
- machine learning
- human activities
- spatio temporal
- video data
- video sequences
- space time
- latent variables
- pairwise
- pedestrian detection
- graph structure
- bayesian networks