Compensating for additive-noise in automatic speech recognition.
Solomon LernerBaruch MazorPublished in: ICSLP (1992)
Keyphrases
- additive noise
- automatic speech recognition
- speech signal
- speech recognition
- hidden markov models
- noisy environments
- noisy images
- word error rate
- spatial domain
- vocal tract
- maximum likelihood
- spoken words
- causal inference
- non stationary
- linear prediction
- broadcast news
- conversational speech
- acoustic features
- random variables
- recognition errors
- word recognition
- speech retrieval
- spontaneous speech
- speech enhancement
- signal processing
- neural network
- speech corpus
- feature extraction
- multiscale
- higher order
- speech synthesis