A new metric for VQ-based speech enhancement and separation.
Mads Græsbøll ChristensenPejman MowlaeePublished in: ICASSP (2011)
Keyphrases
- speech enhancement
- noisy environments
- vector quantization
- noise reduction
- signal to noise ratio
- single channel
- linear prediction
- image coding
- speech signal
- image compression
- vocal tract
- image quality
- background noise
- pattern recognition
- prior information
- image processing
- multi channel
- computational complexity
- wiener filter