A sparse coding approach for local-to-global 3D shape description.
Davide BoscainiUmberto CastellaniPublished in: Vis. Comput. (2014)
Keyphrases
- sparse coding
- dictionary learning
- sparse representation
- natural images
- linear combination
- unsupervised learning
- image classification
- discriminative power
- single image super resolution
- person re identification
- image patches
- group sparsity
- image representation
- feature space
- generative model
- sparse codes
- object category recognition
- high dimensional
- face recognition
- alternating optimization
- spatial pyramid matching
- computer vision
- maximum likelihood
- higher order
- input image
- semi supervised
- probabilistic model
- active learning
- image processing