Visual tracking via incremental Log-Euclidean Riemannian subspace learning.
Xi LiWeiming HuZhongfei ZhangXiaoqin ZhangMingliang ZhuJian ChengPublished in: CVPR (2008)
Keyphrases
- subspace learning algorithm
- image sequences
- visual tracking
- log euclidean
- object tracking
- video sequences
- object appearance
- riemannian metric
- appearance model
- affine invariant
- appearance modeling
- diffusion tensor
- tensor field
- euclidean space
- lie group
- mean shift
- covariance matrices
- vector space
- riemannian manifolds
- particle filter
- partial occlusion
- moving objects
- affine transformation
- subspace learning
- covariance matrix
- appearance variations
- shape analysis
- infinite dimensional
- face recognition
- anisotropic diffusion
- pattern recognition
- pairwise
- semi supervised
- maximum likelihood
- shape space
- geometric structure
- euclidean distance
- shape representation