Object, Scene and Actions: Combining Multiple Features for Human Action Recognition.
Nazli Ikizler-CinbisStan SclaroffPublished in: ECCV (1) (2010)
Keyphrases
- combining multiple
- multiple classifiers
- object features
- multiple objects
- geometric information
- target object
- object models
- object model
- feature set
- complex scenes
- spatial relationships
- spatial relations
- d scene
- image regions
- keypoints
- d objects
- wire frame
- ground plane
- cad model
- geometric constraints
- moving objects
- benchmark datasets
- feature space
- three dimensional
- action recognition
- intensity images
- single image
- fusion methods
- cluster ensemble
- real world scenes
- multiscale