The use of cumulants for voiced-unvoiced segments identification in speech signals.
Baran UsluHakan ToraPublished in: SIU (2014)
Keyphrases
- speech signal
- blind source separation
- speech recognition
- automatic speech recognition
- vocal tract
- speech quality
- hidden markov models
- non stationary
- background noise
- noisy environments
- linear prediction
- formant frequencies
- spectral analysis
- speaker identification
- speech enhancement
- automatic speech recognition systems
- additive noise
- fourth order
- principal component analysis
- multiresolution
- machine learning