On microphone arrangement for multichannel speech enhancement based on nonnegative matrix factorization in time-channel domain.
Yoshikazu MuraseHironobu ChibaNobutaka OnoShigeki MiyabeTakeshi YamadaShoji MakinoPublished in: APSIPA (2014)
Keyphrases
- single channel
- nonnegative matrix factorization
- speech enhancement
- multi channel
- sound source
- matrix factorization
- negative matrix factorization
- least squares
- data representation
- independent component analysis
- spectral clustering
- linear prediction
- frequency domain
- signal processing
- principal component analysis
- noise reduction
- prior information
- noisy environments
- wiener filter
- data sets
- speech signal
- wavelet decomposition
- high dimensional