High-likelihood model based on reliability statistics for robust combination of features: application to noisy speech recognition.
Peter JancovicMünevver KöküerFionn MurtaghPublished in: INTERSPEECH (2003)
Keyphrases
- speech recognition
- noisy environments
- speech recognition systems
- hidden markov models
- speech signal
- cepstral coefficients
- language model
- speech processing
- automatic speech recognition
- speech recognizer
- feature set
- speaker identification
- pattern recognition
- keyword spotting
- speech synthesis
- speech recognition technology
- maximum likelihood
- low level
- feature extraction
- computer vision
- structural features
- extracted features
- key frames
- speech understanding
- feature vectors
- speaker independent
- neural network