Data-driven spatio-temporal RGBD feature encoding for action recognition in operating rooms.
Andru Putra TwinandaEmre O. AlkanAfshin GangiMichel de MathelinNicolas PadoyPublished in: Int. J. Comput. Assist. Radiol. Surg. (2015)
Keyphrases
- action recognition
- data driven
- spatio temporal interest points
- human actions
- spatio temporal
- depth sensors
- spatial temporal
- action recognition in videos
- view invariant
- bag of words
- action classification
- motion features
- activity recognition
- recognition of human actions
- human detection
- computer vision
- recognizing human actions
- body parts
- spatial and temporal
- human pose
- recognizing actions
- feature vectors
- action primitives
- human motion
- human activities
- view invariant action recognition
- foreground segmentation
- global features
- static images
- machine learning
- human object interactions
- image features