Modeling with a subspace constraint on inverse covariance matrices.
Scott AxelrodRamesh GopinathPeder A. OlsenPublished in: INTERSPEECH (2002)
Keyphrases
- covariance matrices
- covariance matrix
- maximum likelihood
- vector space
- gaussian distribution
- distance measure
- gaussian mixture model
- multivariate normal
- feature vectors
- principal component analysis
- gaussian mixture
- transformation matrix
- riemannian manifolds
- linear classifiers
- similarity measure
- learning algorithm
- log euclidean
- riemannian metric
- image processing
- image segmentation
- feature space
- lower bound
- high dimensional data
- high dimensional
- low dimensional
- image classification
- dimensionality reduction