Combining Spatio-Temporal Appearance Descriptors and Optical Flow for Human Action Recognition in Video Data.
Karla BrkicSrdan RasicAxel PinzSinisa SegvicZoran KalafaticPublished in: CoRR (2013)
Keyphrases
- video data
- spatio temporal
- optical flow
- video sequences
- image sequences
- appearance descriptors
- moving objects
- human actions
- motion features
- video streams
- video analysis
- motion trajectories
- bag of features
- action recognition
- temporal structure
- video content
- multimedia
- video database
- video camera
- video frames
- video indexing
- key frames
- space time
- flow field
- video retrieval
- motion estimation
- interesting events
- motion segmentation
- motion field
- motion model
- computer vision
- motion patterns
- image processing
- machine learning