EXMOVES: Classifier-based Features for Scalable Action Recognition.
Du TranLorenzo TorresaniPublished in: ICLR (Poster) (2014)
Keyphrases
- action recognition
- bag of features
- action classification
- motion features
- static images
- spatio temporal interest points
- human detection
- computer vision
- human actions
- bag of words
- spatial temporal
- activity recognition
- mid level
- human movements
- co occurrence
- independent subspace analysis
- recognition of human actions
- recognizing actions
- feature vectors
- action primitives
- space time interest points
- depth cameras
- salient features
- body parts
- feature space
- depth sensors
- deformable part models
- motion history images
- global features
- recognizing human actions
- feature detection
- low level