Motion Feature Combination for Human Action Recognition in Video.
Hongying MengNick E. PearsChris BaileyPublished in: VISIGRAPP (Selected Papers) (2007)
Keyphrases
- motion features
- space time
- action recognition
- human actions
- video sequences
- video data
- spatial and temporal
- object motion
- motion analysis
- temporal filtering
- input video
- motion patterns
- key frames
- moving camera
- image sequences
- video frames
- motion estimation
- motion model
- successive frames
- visual cues
- visual data
- dynamic scenes
- temporal consistency
- temporal continuity
- video footage
- temporal domain
- video representation
- low frame rate
- visual features
- camera motion
- activity recognition
- motion trajectories
- motion parameters
- temporal coherence
- video scene
- surveillance videos
- single frame
- feature vectors
- image frames
- human motion
- video streams
- spatio temporal
- multimedia
- layered representation
- shot change detection
- image features
- video shots
- video content
- motion field
- motion detection
- video retrieval
- video surveillance
- object tracking
- static images
- periodic motion
- multi view
- multiview video
- stationary camera
- feature set
- motion capture data
- reference frame
- video objects