How to avoid vowel normalization in identification of vowels in continuous speech.
Maria-Gabriella Di BenedettoArmando LanaroPublished in: ICASSP (1984)
Keyphrases
- formant frequencies
- vocal tract
- speech recognition systems
- speech signal
- speech recognition
- speech synthesis
- vowel phonemes
- cross correlation
- hidden markov models
- image acquisition
- normalization method
- identification rate
- data sets
- template matching
- preprocessing
- person identification
- english text
- search algorithm
- artificial intelligence
- learning algorithm
- genetic algorithm