Music Signal Separation Based on Supervised Nonnegative Matrix Factorization with Orthogonality and Maximum-Divergence Penalties.
Daichi KitamuraHiroshi SaruwatariKosuke YagiKiyohiro ShikanoYu TakahashiKazunobu KondoPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2014)
Keyphrases
- nonnegative matrix factorization
- probabilistic latent semantic indexing
- data representation
- negative matrix factorization
- matrix factorization
- least squares
- signal processing
- principal component analysis
- nonnegative matrix
- semi supervised
- spectral clustering
- document clustering
- original data
- unsupervised learning
- learning algorithm
- machine learning
- natural language processing
- pairwise
- image processing
- sparsity constraints
- computer vision