Non-negative matrix factorization for highly noise-robust ASR: To enhance or to recognize?
Felix WeningerMartin WöllmerJürgen T. GeigerBjörn W. SchullerJort F. GemmekeAntti HurmalainenTuomas VirtanenGerhard RigollPublished in: ICASSP (2012)
Keyphrases
- negative matrix factorization
- noisy environments
- sparsity constraints
- sparse representation
- spectral clustering
- matrix factorization
- nonnegative matrix factorization
- document clustering
- latent semantic space
- constrained least squares
- speech recognition
- principal component analysis
- automatic speech recognition
- tensor factorization
- feature vectors
- clustering algorithm
- data mining