Speech enhancement based on a few shapes of speech spectrum.
Feng BaoHui-jing DouMao-shen JiaChangchun BaoPublished in: ChinaSIP (2014)
Keyphrases
- speech enhancement
- noisy environments
- speech signal
- spectral subtraction
- noise reduction
- signal to noise ratio
- single channel
- vocal tract
- linear prediction
- noisy speech
- speech recognition
- wiener filter
- smoothing algorithm
- background noise
- additive noise
- hidden markov models
- automatic speech recognition
- speaker identification
- speaker recognition
- multi channel
- speech synthesis
- non stationary
- image restoration