Asymptotically exact noise-corrupted speech likelihoods.
Rogier C. van DalenMark J. F. GalesPublished in: INTERSPEECH (2010)
Keyphrases
- speech recognition
- audio visual
- speech signal
- speaker recognition
- sample size
- endpoint detection
- text to speech
- posterior probability
- random variables
- maximum likelihood
- worst case
- brute force
- dialogue system
- likelihood function
- noisy environments
- speech processing
- upper bound
- automatic speech recognition
- hidden markov models
- neural network
- recognition engine
- data sets