Do We Really Have to Consider Covariance Matrices for Image Features?.
Yasushi KanazawaKen-ichi KanataniPublished in: ICCV (2001)
Keyphrases
- covariance matrices
- image features
- covariance matrix
- feature vectors
- pixel intensities
- maximum likelihood
- vector space
- gaussian mixture model
- distance measure
- gaussian distribution
- image classification
- multivariate normal
- computer vision
- object recognition
- riemannian metric
- principal component analysis
- gaussian mixture
- riemannian manifolds
- query processing
- feature space
- optimal solution
- similarity measure