Parametric-Pearson-based independent component analysis for frequency-domain blind speech separation.
Hiroko KatoYuichi NagaharaShoko ArakiHiroshi SawadaShoji MakinoPublished in: EUSIPCO (2006)
Keyphrases
- frequency domain
- independent component analysis
- source signals
- spatial domain
- blind separation
- principal component analysis
- independent components
- signal processing
- blind source separation
- factor analysis
- feature extraction
- speech signal
- speech recognition
- power spectrum
- denoising
- audio visual
- subband
- single channel
- low pass
- source separation
- machine learning
- hidden markov models
- image data