Semantic Retrieval for Videos in Non-static Background Using Motion Saliency and Global Features.
Dianting LiuMei-Ling ShyuPublished in: ICSC (2013)
Keyphrases
- static background
- semantic retrieval
- global features
- static camera
- motion detection
- moving camera
- video surveillance
- moving objects
- dynamic scenes
- image sequences
- background subtraction
- visual features
- foreground objects
- video sequences
- image annotation
- object motion
- feature vectors
- background model
- motion model
- human actions
- keypoints
- camera motion
- saliency map
- image retrieval
- optical flow
- video frames
- motion estimation
- space time
- image features
- detection algorithm
- classification method
- key frames
- motion segmentation
- human motion
- video data
- spatio temporal
- object tracking
- low level
- three dimensional
- visual attention
- learning algorithm
- computer vision
- search engine