A visual attention model for dynamic scenes based on motion features.
Changle ZhouJiawei ChenJinliang YaoPublished in: ICARCV (2012)
Keyphrases
- dynamic scenes
- motion features
- space time
- motion patterns
- video sequences
- video data
- motion analysis
- activity recognition
- saliency map
- multi view
- moving objects
- visual features
- motion vectors
- motion segmentation
- action recognition
- optical flow
- video shots
- visual attention
- human actions
- motion model
- flow field
- background subtraction
- image sequences
- video frames
- spatio temporal
- frame rate
- video streams
- eye movements
- machine learning
- eye tracking
- low level features
- model selection
- video retrieval
- image classification
- object detection
- motion estimation
- computational complexity