Decomposition of the speech signal into short-time waveforms using spectral segmentation.
Christophe d'AlessandroJean-Sylvain LiénardPublished in: ICASSP (1988)
Keyphrases
- speech signal
- spectral analysis
- linear prediction
- speech recognition
- spectral subtraction
- cepstral coefficients
- non stationary
- speaker identification
- multiscale
- automatic speech recognition
- noisy environments
- image segmentation
- vocal tract
- hidden markov models
- background noise
- speaker verification
- speaker recognition
- acoustic features
- speech quality
- automatic speech recognition systems
- prediction error
- speech enhancement
- fundamental frequency
- multiresolution
- machine learning