Two-stage sparse coding of region covariance via Log-Euclidean kernels to detect saliency.
Ying-Ying ZhangCai YangPing ZhangPublished in: Neural Networks (2017)
Keyphrases
- sparse coding
- linear combination
- covariance matrices
- log euclidean
- dictionary learning
- natural images
- feature space
- sparse representation
- covariance matrix
- image classification
- unsupervised learning
- image representation
- riemannian manifolds
- generative model
- image patches
- input image
- euclidean space
- riemannian metric
- visual words
- maximum likelihood
- high dimensional
- kernel function
- affine invariant
- dimensionality reduction
- vector space
- higher order
- probabilistic model
- gaussian distribution
- geometric structure
- feature vectors
- kernel methods
- kernel matrix
- tensor field
- lie group
- object recognition