Performance evaluation of nonlinear speech enhancement based on virtual increase of channels in reverberant environments.
Kouei YamaokaShoji MakinoNobutaka OnoTakeshi YamadaPublished in: EUSIPCO (2017)
Keyphrases
- speech enhancement
- speech signal
- single channel
- noise reduction
- noisy environments
- multi channel
- speech recognition
- virtual environment
- augmented reality
- additive noise
- non stationary
- linear prediction
- signal to noise ratio
- automatic speech recognition
- adaptive filtering
- image processing
- probabilistic model
- vocal tract
- real environment
- background noise
- dynamic environments
- wiener filter
- edge preserving
- sound source
- low contrast
- frequency domain
- edge detection
- multiscale