An Investigation into the Correlation and Prediction of Acoustic Speech Features from MFCC Vectors.
Jonathan DarchBen MilnerIbrahim AlmajaiSaeed VaseghiPublished in: ICASSP (4) (2007)
Keyphrases
- feature vectors
- mel frequency cepstral coefficients
- speech recognition systems
- speech recognition
- speech signal
- feature extraction
- speaker recognition
- speaker identification
- feature set
- acoustic features
- low level
- prediction accuracy
- audio visual
- audio features
- speech segments
- automatic speech recognition
- emotional speech
- linear predictive
- linear prediction
- non stationary
- extracted features
- prediction error
- speaker diarization
- classification accuracy
- prediction model
- cepstral coefficients
- gaussian mixture model