Fast distributed multichannel speech enhancement using novel frequency domain estimators of magnitude-squared spectrum.
Jingxian TuYoushen XiaPublished in: Speech Commun. (2015)
Keyphrases
- frequency domain
- single channel
- speech enhancement
- power spectrum
- spatial domain
- fourier transform
- multi channel
- frequency spectrum
- denoising
- power spectral
- subband
- feature extraction
- additive noise
- linear prediction
- least squares
- signal to noise ratio
- noise reduction
- prior information
- wavelet domain
- independent component analysis
- wavelet decomposition
- noisy environments
- pattern recognition
- machine learning