Riemannian submanifold framework for log-Euclidean metric learning on symmetric positive definite manifolds.
Sung Woo ParkJunseok KwonPublished in: Expert Syst. Appl. (2022)
Keyphrases
- riemannian manifolds
- log euclidean
- euclidean space
- spd matrices
- symmetric positive definite
- riemannian metric
- diffusion tensor
- metric learning
- lie group
- vector space
- covariance matrices
- distance metric
- tensor field
- metric space
- multi class classification
- euclidean distance
- manifold learning
- fisher information
- covariance matrix
- geometric structure
- feature space
- affine invariant
- mean shift
- low dimensional
- pairwise
- geodesic distance
- shape analysis
- distance function
- multi task
- parameter space
- data points