Compensation of speaker directivity in speech recognition using HMM composition.
Franck GironYasuhiro MinamiMasashi TanakaKen'ichi FuruyaPublished in: ICASSP (1998)
Keyphrases
- speech recognition
- automatic speech recognition
- automatic speech recognition systems
- speech signal
- hidden markov models
- speaker adaptation
- speaker independent
- speaker identification
- speaker dependent
- vocal tract
- speaker recognition
- acoustic models
- sound source
- speech recognizer
- speech synthesis
- speech processing
- visual speech
- audio visual
- phoneme recognition
- keyword spotting
- multi stream
- pattern recognition
- noisy speech
- speaker verification
- acoustic features
- noisy environments
- web service composition
- language model
- broadcast news
- linear prediction
- recognition engine
- speech recognition systems
- prosodic features
- em algorithm
- mel frequency cepstral coefficients
- visual speech recognition
- speech sounds
- synthesized speech
- handwriting recognition
- finite state transducers
- audio stream
- spontaneous speech
- text to speech