Combining sparse and dense descriptors with temporal semantic structures for robust human action recognition.
Jie ChenGuoying ZhaoVili-Petteri KellokumpuMatti PietikäinenPublished in: ICCV Workshops (2011)
Keyphrases
- metadata
- action recognition
- semantic structures
- motion history images
- action recognition in videos
- bag of features
- static images
- human actions
- view invariant
- bag of words
- computer vision
- human activities
- human detection
- action classification
- activity recognition
- human object interactions
- feature descriptors
- mid level features
- body parts
- human pose
- spatio temporal
- semantic network
- recognition of human actions
- pose estimation
- recognizing human actions
- human motion
- low level descriptors
- atomic actions
- knowledge domains
- d objects
- space time
- temporal structure
- image classification