A Canonical Correlation Analysis based motion model for probabilistic visual tracking.
Tom HeymanVincent SpruytSebastian GruenwedelAlessandro LeddaWilfried PhilipsPublished in: VCIP (2012)
Keyphrases
- visual tracking
- motion model
- canonical correlation analysis
- particle filter
- data association
- particle filtering
- observation model
- object tracking
- appearance model
- motion analysis
- mean shift
- motion patterns
- motion segmentation
- kalman filter
- motion estimation
- state space
- partial least squares
- generative model
- robust tracking
- image sequences
- proposal distribution
- video sequences
- optical flow
- least squares
- probabilistic model
- dynamical models
- image processing