A noise-robust speech recognition approach incorporating normalized speech/non-speech likelihood into hypothesis scores.
Tasuku OonishiKoji IwanoSadaoki FuruiPublished in: Speech Commun. (2013)
Keyphrases
- speech recognition
- noisy environments
- speech signal
- speech synthesis
- automatic speech recognition
- spectral subtraction
- speech recognizer
- noisy speech
- speech processing
- speaker identification
- hidden markov models
- speech recognition technology
- language model
- pattern recognition
- speech enhancement
- background noise
- recognition engine
- speaker verification
- speech recognition systems
- word error rate
- handwriting recognition
- speaker independent
- speech recognizers
- speech recognition errors
- vocal tract
- information retrieval
- voice activity detection
- speaker recognition
- isolated word
- maximum likelihood
- acoustic models
- signal processing
- cepstral coefficients
- speaker adaptation
- signal to noise ratio