MAP prediction of formant frequencies and voicing class from MFCC vectors in noise.
Jonathan DarchBen P. MilnerSaeed VaseghiPublished in: Speech Commun. (2006)
Keyphrases
- prediction accuracy
- feature vectors
- speech signal
- prediction algorithm
- noise reduction
- maximum a posteriori
- formant frequencies
- prediction model
- signal to noise ratio
- missing data
- pattern recognition
- feature extraction
- speech recognition
- image reconstruction
- vector space
- vector quantization
- maximum likelihood
- edge detection
- median filter
- noise model
- noisy environments
- background noise
- input pattern
- input vector
- machine learning