Detecting abnormal events in video using Narrowed Motion Clusters.
Radu Tudor IonescuSorina SmeureanuMarius PopescuBogdan AlexePublished in: CoRR (2018)
Keyphrases
- space time
- surveillance videos
- motion trajectories
- video sequences
- event detection
- event recognition
- spatial and temporal
- video clips
- video data
- temporal structure
- motion features
- input video
- object motion
- temporal filtering
- video event
- video frames
- motion estimation
- human activities
- image sequences
- key frames
- moving camera
- motion analysis
- temporal continuity
- clustering algorithm
- video streams
- dynamic scenes
- successive frames
- dynamic textures
- motion patterns
- visual cues
- low frame rate
- static images
- video analysis
- video content
- video scene
- temporal consistency
- moving objects
- visual data
- shot change detection
- video objects
- temporal information
- video footage
- motion field
- video summarization
- sports video
- motion model
- camera motion
- temporal relationships
- temporal coherence
- optical flow
- spatio temporal
- data points
- layered representation
- video surveillance
- news stories
- human motion
- temporal relations
- activity recognition
- multimedia
- video retrieval
- soccer video
- image motion
- motion capture data
- temporal events
- image frames
- reference frame
- humanoid robot
- interesting events
- motion segmentation