MSR-CNN: Applying motion salient region based descriptors for action recognition.
Zhigang TuJun CaoYikang LiBaoxin LiPublished in: ICPR (2016)
Keyphrases
- action recognition
- action recognition in videos
- recognizing human actions
- human actions
- static images
- human movements
- recognition of human actions
- bag of words
- bag of features
- motion features
- human detection
- activity recognition
- computer vision
- view invariant
- action classification
- action primitives
- body parts
- mid level
- motion capture data
- motion history images
- recognizing actions
- image retrieval
- depth sensors
- image segmentation
- mid level features
- salient features
- histogram of oriented gradients
- human pose
- visual words
- human activities
- human computer interaction
- image sequences
- space time interest points