Spatio-temporal multi-scale motion descriptor from a spatially-constrained decomposition for online action recognition.
Fabio MartínezAntoine ManzaneraEduardo RomeroPublished in: IET Comput. Vis. (2017)
Keyphrases
- action recognition
- action recognition in videos
- recognition of human actions
- human actions
- multiscale
- view invariant
- spatio temporal
- recognizing human actions
- static images
- motion features
- human movements
- bag of words
- spatio temporal interest points
- activity recognition
- spatial temporal
- action primitives
- human detection
- computer vision
- action classification
- space time
- recognizing actions
- motion patterns
- image sequences
- spatial and temporal
- body parts
- mid level
- motion recognition
- human motion
- moving objects
- motion estimation
- motion analysis
- human pose
- motion capture data
- action detection
- visual words
- motion history images
- human activities
- bag of features
- motion model
- keypoints
- image representation
- shape descriptors
- optical flow
- motion trajectories