A novel target speaker dependent postfiltering approach for multichannel speech enhancement.
Ritwik GiriKarim HelwaniTao ZhangPublished in: WASPAA (2017)
Keyphrases
- speech enhancement
- single channel
- speaker dependent
- speech recognition
- noisy environments
- multi channel
- speaker identification
- linear prediction
- speech signal
- noise reduction
- signal to noise ratio
- vocal tract
- independent component analysis
- image coding
- image segmentation
- pattern recognition
- speaker independent
- brain activity
- wiener filter
- wavelet decomposition
- frequency domain
- non stationary