Jointly Optimizing Activation Coefficients of Convolutive NMF Using DNN for Speech Separation.
Hao LiShuai NieXueliang ZhangHui ZhangPublished in: INTERSPEECH (2016)
Keyphrases
- negative matrix factorization
- nonnegative matrix factorization
- linear prediction
- sparsity constraints
- speech recognition
- blind separation
- linear combination
- basis vectors
- speech synthesis
- linear predictive coding
- speech signal
- wavelet coefficients
- basis functions
- automatic speech recognition
- blind source separation
- recognition engine
- matrix factorisation
- audio visual
- information processing
- training process
- matrix factorization
- spoken language
- cepstral coefficients
- broadcast news
- text to speech
- spectral clustering
- data sets
- speaker identification
- sparse coding
- document clustering
- sparse representation
- hidden markov models