Using missing feature theory to actively select features for robust speech recognition with interruptions, filtering and noise KN-37.
Richard P. LippmannBeth A. CarlsonPublished in: EUROSPEECH (1997)
Keyphrases
- speech recognition
- noisy environments
- feature set
- cepstral coefficients
- speech recognition systems
- feature vectors
- image features
- automatic speech recognition
- speech signal
- salt pepper
- language model
- noise reduction
- speech synthesis
- hidden markov models
- background noise
- speaker identification
- noisy speech
- pattern recognition
- mel frequency cepstral coefficients
- speech processing
- speech recognition technology
- speaker dependent
- speaker independent
- classification accuracy
- feature space
- visual speech
- feature extraction
- machine learning
- low level
- signal processing
- neural network