Prediction of Fundamental Frequency and Voicing From Mel-Frequency Cepstral Coefficients for Unconstrained Speech Reconstruction.
Ben MilnerXu ShaoPublished in: IEEE Trans. Speech Audio Process. (2007)
Keyphrases
- speech signal
- fundamental frequency
- mel frequency cepstral coefficients
- linear prediction
- speech recognition
- speaker identification
- automatic speech recognition
- speaker recognition
- acoustic features
- prediction accuracy
- noisy environments
- hidden markov models
- cepstral coefficients
- linear predictive
- non stationary
- spectral analysis
- noisy images
- feature set
- pattern recognition
- image processing
- probabilistic neural network
- language model
- speaker diarization
- visual speech
- machine learning