Ordered Pooling of Optical Flow Sequences for Action Recognition.
Jue WangAnoop CherianFatih PorikliPublished in: CoRR (2017)
Keyphrases
- action recognition
- optical flow
- computer vision
- motion features
- human actions
- bag of words
- action classification
- activity recognition
- human detection
- motion capture data
- spatial temporal
- hidden markov models
- motion history images
- image sequences
- view invariant
- motion field
- recognizing human actions
- bag of features
- mid level
- body parts
- d scene
- human activities
- recognizing actions
- depth sensors
- machine learning
- camera motion
- space time interest points
- recognition of human actions
- action detection
- scene flow
- motion model
- depth cameras
- pose estimation
- spatio temporal
- moving objects
- video dataset
- max margin
- human pose
- view invariant action recognition