An EM-algorithm approach for the design of orthonormal bases adapted to sparse representations.
Angélique DremeauCédric HerzetPublished in: ICASSP (2010)
Keyphrases
- em algorithm
- sparse representation
- expectation maximization
- maximum likelihood
- mixture model
- gaussian mixture model
- generative model
- parameter estimation
- maximum likelihood estimation
- expectation maximisation
- machine learning
- probability density function
- maximum a posteriori
- dictionary learning
- dimensionality reduction
- image classification
- sparse coding
- hyperparameters
- gaussian mixture
- parameter learning
- sparse coefficients
- manifold learning
- linear combination
- data analysis
- pattern recognition
- image processing