Autonomous Visual Events Detection and Classification without Explicit Object-Centred Segmentation and Tracking.
Tao XiangShaogang GongDennis ParkinsonPublished in: BMVC (2002)
Keyphrases
- bounding box
- multiple objects
- cluttered background
- video scene
- features extraction
- detection responses
- feature extraction and classification
- detecting objects
- partial occlusion
- deformable objects
- object tracking
- object detection and tracking
- object detection
- motion segmentation
- object segmentation
- event detection
- d objects
- moving objects
- motion cues
- occlusion reasoning
- articulated objects
- complex background
- feature extraction
- target object
- discriminative classifiers
- foreground background separation
- image classification
- segmentation method
- appearance model
- detecting and tracking multiple
- detection algorithm
- human detection
- spatial relations
- motion model
- image regions
- visual information
- segmentation algorithm
- visual features
- foreground regions
- graph cuts
- particle filter
- surveillance videos
- input image
- feature vectors
- accurate localization
- feature space
- image segmentation