Evaluating spatiotemporal interest point features for depth-based action recognition.
Yu ZhuWenbin ChenGuodong GuoPublished in: Image Vis. Comput. (2014)
Keyphrases
- action recognition
- point features
- structure from motion
- multiple views
- computer vision
- point correspondences
- image features
- spatial and temporal
- feature points
- body parts
- data association
- depth map
- spatio temporal
- space time
- line segments
- viewpoint
- depth information
- projective transformations
- single image
- d objects
- moving objects
- motion parameters
- video sequences
- depth images
- feature extraction