A hybrid speech enhancement system based on HMM and spectral subtraction.
M. H. GhoreishiHamid SheikhzadehPublished in: ICASSP (2000)
Keyphrases
- spectral subtraction
- speech signal
- speech enhancement
- speech recognition
- noisy speech
- hidden markov models
- automatic speech recognition
- noisy environments
- noise reduction
- linear prediction
- signal to noise ratio
- background noise
- vocal tract
- single channel
- additive noise
- speech synthesis
- adaptive filtering
- non stationary
- wiener filter
- hearing aids
- image restoration