Speaker-independent 100 word recognition using dynamic spectral features of speech and a neural network.
Tadashi KitamuraPublished in: EUROSPEECH (1993)
Keyphrases
- spectral features
- speech recognition
- speaker independent
- speech recognizer
- speech recognition systems
- neural network
- speaker dependent
- remote sensing images
- spectral analysis
- structural features
- hyperspectral images
- speaker identification
- spoken language
- hidden markov models
- digit recognition
- mel frequency cepstral coefficients
- speaker adaptation
- pattern recognition
- speech signal
- acoustic models
- broadcast news
- remote sensing
- speaker recognition
- language model
- co occurrence
- image data