Removal of residual cross-talk components in Blind Source Separation using time-delayed spectral subtraction.
Ryo MukaiShoko ArakiHiroshi SawadaShoji MakinoPublished in: ICASSP (2002)
Keyphrases
- blind source separation
- speech signal
- independent components
- independent component analysis
- source separation
- non stationary
- blind separation
- negative matrix factorization
- speech recognition
- face recognition
- principal component analysis
- factor analysis
- training set
- hidden markov models
- automatic speech recognition