Transformations of speech spectra to a two-dimensional continuous-valued phonetic feature space for vowel training.
Andrew E. BeckStephen A. ZahorianPublished in: ICASSP (1992)
Keyphrases
- continuous valued
- speech recognition
- feature space
- speech signal
- speaker independent
- speech synthesis
- acoustic models
- speech recognition systems
- training set
- training samples
- principal component analysis
- vocal tract
- hidden markov models
- discrete valued
- speech recognizer
- formant frequencies
- hearing impaired
- automatic speech recognition
- speaker identification
- spoken document retrieval
- high dimensional
- fundamental frequency
- spectral features
- pattern recognition
- speech sounds
- feature extraction
- text to speech
- spoken term detection
- prosodic features
- non stationary
- feature selection
- data points
- language model
- kernel function
- multiresolution
- supervised learning