Total Variation and Signature-Based Regularizations on Coupled Nonnegative Matrix Factorization for Data Fusion.
Feixia YangFei MaZiliang PingGuixian XuPublished in: IEEE Access (2019)
Keyphrases
- data fusion
- nonnegative matrix factorization
- total variation
- minimization problems
- interior point
- image restoration
- regularization term
- denoising
- image denoising
- matrix factorization
- negative matrix factorization
- data representation
- convex optimization
- least squares
- principal component analysis
- augmented lagrangian
- image processing
- spectral clustering
- low rank
- document clustering
- original data
- natural images
- linear programming
- markov random field
- multiresolution
- objective function
- similarity measure