Tracking Humans using Prior and Learned Representations of Shape and Appearance.
Jongwoo LimDavid J. KriegmanPublished in: FGR (2004)
Keyphrases
- appearance features
- appearance manifolds
- appearance model
- appearance information
- shape change
- appearance variations
- object appearance
- pose recovery
- shape constraints
- appearance modeling
- visual tracking
- human faces
- object tracking
- object shape
- probabilistic principal component analysis
- target object
- pedestrian detection and tracking
- likelihood model
- object shapes
- shape model
- appearance cues
- particle filter
- particle filtering
- deformable objects
- prior information
- shape representation
- object representation
- kalman filter
- face recognition
- pose estimation
- facial expressions
- camera views
- shape features
- background subtraction
- prior knowledge
- visual input
- spatial layout
- set of training images
- articulated objects
- object class
- shape descriptors
- deformable models
- action recognition
- graph cuts
- dynamically updated
- detection responses
- articulated human motion
- video sequences