Mask classification for missing-feature reconstruction for robust speech recognition in unknown background noise.
Wooil KimRichard M. SternPublished in: Speech Commun. (2011)
Keyphrases
- speech recognition
- noisy environments
- background noise
- speech signal
- pattern recognition
- digit recognition
- noisy speech
- hidden markov models
- language model
- feature vectors
- cepstral coefficients
- automatic speech recognition
- speaker identification
- speech recognition technology
- feature set
- speech enhancement
- noise reduction
- speech recognizer
- text classification
- feature extraction
- feature selection
- speech synthesis
- speaker independent
- speech recognition systems
- vocal tract
- image classification
- discriminative training
- classification accuracy
- spectral subtraction
- machine learning
- speaker dependent