BSIK-SVD: A dictionary-learning algorithm for block-sparse representations.
Yongqin ZhangJiaying LiuMading LiZongming GuoPublished in: ICASSP (2014)
Keyphrases
- sparse representation
- learning algorithm
- singular value decomposition
- reconstruction error
- dimensionality reduction
- dictionary learning
- sparse coding
- face recognition
- learned dictionary
- random projections
- image patches
- active learning
- image classification
- compressive sensing
- training data
- joint optimization
- high dimensional data
- regularized least squares
- compressed sensing
- negative matrix factorization
- signal processing
- image representation
- machine learning
- least squares
- sparse reconstruction
- neural network
- learning problems
- singular values
- image super resolution
- sparse coefficients
- multi task learning
- principal component analysis
- test images
- orthogonal matching pursuit