Adaptive noise power estimation using spectral difference for robust speech enhancement.
Jae-Hun ChoiSang-Kyun KimJoon-Hyuk ChangPublished in: ICASSP (2012)
Keyphrases
- speech enhancement
- noisy environments
- noise reduction
- spectral subtraction
- linear prediction
- signal to noise ratio
- background noise
- speech signal
- single channel
- speech recognition
- noisy speech
- wiener filter
- multi channel
- spectral analysis
- prediction error
- image restoration
- automatic speech recognition
- edge preserving