Is the vowel length contrast in Japanese exaggerated in infant-directed speech?
Keiichi TajimaKuniyoshi TanakaAndrew MartinReiko MazukaPublished in: INTERSPEECH (2013)
Keyphrases
- speech signal
- speech synthesis
- vocal tract
- formant frequencies
- speech recognition
- prosodic features
- speech recognition systems
- fundamental frequency
- text to speech
- automatic speech recognition
- speaker verification
- total length
- speaker independent
- noisy environments
- speaker recognition
- hidden markov models
- endpoint detection
- automatic speech recognition systems
- real time
- vowel phonemes
- speaker identification
- spoken language
- maximum number
- study proposes
- speech recognizer
- fixed length
- image processing
- genetic algorithm
- neural network