A unified model of appearance and motion of video and its application in STIP detection.
Yanshan LiRongjie XiaWeixin XiePublished in: Signal Image Video Process. (2018)
Keyphrases
- unified model
- motion features
- object motion
- bounding box
- motion cues
- static images
- object detection and tracking
- space time
- shot change detection
- dynamic textures
- video scene
- motion analysis
- video sequences
- video data
- appearance features
- motion patterns
- action recognition
- spatial and temporal
- temporal segmentation
- motion model
- activity recognition
- layered representation
- video shots
- visual data
- image sequences
- dynamic scenes
- camera motion
- optical flow
- temporal filtering
- face detection and tracking
- motion estimation
- input video
- moving camera
- video streams
- video recordings
- detection algorithm
- video analysis
- human motion
- moving objects
- multimedia
- key frames
- video frames
- video objects
- multi channel
- object detection
- event detection
- video surveillance
- motion vectors
- foreground regions
- spatio temporal
- object tracking
- articulated human motion
- shot detection
- depth images
- color features
- appearance model
- human actions
- video clips
- motion segmentation