Robust speech recognition training via duration and spectral-based stress token generation.
John H. L. HansenSahar E. Bou-GhazalePublished in: IEEE Trans. Speech Audio Process. (1995)
Keyphrases
- speech recognition
- wall street journal corpus
- noisy environments
- hidden markov models
- speech signal
- isolated word
- cepstral coefficients
- automatic speech recognition
- speech synthesis
- speech recognition technology
- speech recognition systems
- speech recognizer
- speech processing
- language model
- acoustic models
- speech understanding
- speech retrieval
- speaker dependent
- speaker adaptation
- discriminative training
- keyword spotting
- speaker independent
- training process
- speech recognition errors
- speaker identification
- pattern recognition
- remote sensing
- speaker diarization
- machine learning
- maximum likelihood
- speech recognizers