Unsupervised Learning on Monocular Videos for 3D Human Pose Estimation.
Sina HonariVictor ConstantinHelge RhodinMathieu SalzmannPascal FuaPublished in: CoRR (2020)
Keyphrases
- unsupervised learning
- view invariant
- video sequences
- image sequences
- action detection
- pose estimation
- supervised learning
- action recognition
- body parts
- semi supervised
- deep learning
- video content
- video data
- video database
- video surveillance
- independently moving objects
- feature selection
- dimensionality reduction
- machine learning
- model selection
- video frames
- human activities
- active learning
- stereo camera
- object recognition
- key frames
- text classification
- human actions
- optical flow
- spatio temporal
- moving objects
- deep architectures
- model based tracking