Speech recognition using voice-characteristic-dependent acoustic models.
Hiroyuki SuzukiHeiga ZenYoshihiko NankakuChiyomi MiyajimaKeiichi TokudaTadashi KitamuraPublished in: ICASSP (1) (2003)
Keyphrases
- acoustic models
- speech recognition
- speech recognizer
- automatic speech recognition
- broadcast news
- hidden markov models
- speaker independent
- text to speech
- speech signal
- speech sounds
- spoken language
- discriminative training
- training data
- emotion recognition
- speaker identification
- fundamental frequency
- speech quality
- language model
- mel frequency cepstral coefficients
- speech recognition systems
- pattern recognition
- information retrieval
- non stationary
- probability distribution
- image processing