Making Action Recognition Robust to Occlusions and Viewpoint Changes.
Daniel WeinlandMustafa ÖzuysalPascal FuaPublished in: ECCV (3) (2010)
Keyphrases
- action recognition
- view invariant
- viewpoint
- computer vision
- human actions
- bag of words
- human detection
- activity recognition
- multi view
- robust tracking
- action classification
- body parts
- single view
- spatial temporal
- partial occlusion
- bag of features
- depth sensors
- recognizing human actions
- image sequences
- action recognition in videos
- multiple viewpoints
- recognizing actions
- multiple views
- recognition of human actions
- independent subspace analysis
- view invariant action recognition
- video dataset
- static images
- action primitives
- object tracking
- object detection
- spatio temporal