Subspace tracking based on the constrained projection approximation approach.
Amir ValizadehMahmood KarimiPublished in: EUSIPCO (2007)
Keyphrases
- visual tracking
- particle filter
- linear projection
- approximation algorithms
- low dimensional
- sequential monte carlo methods
- real time
- projection matrix
- kalman filter
- null space
- feature space
- feature selection
- high dimensional data
- motion analysis
- dimensionality reduction
- feature extraction
- partial occlusion
- computer vision
- eigendecomposition
- approximation error
- subspace methods
- subspace learning
- convex sets
- motion segmentation
- error bounds
- closed form
- appearance model
- human body
- high dimensional
- principal component analysis