Sparse smoothing of articulatory features from Gaussian mixture model based acoustic-to-articulatory inversion: benefit to speech recognition.
Prasad SudhakarPrasanta Kumar GhoshPublished in: INTERSPEECH (2014)
Keyphrases
- speech recognition
- speech recognition systems
- language model
- speaker independent
- speech signal
- hidden markov models
- speech synthesis
- speech recognizers
- automatic speech recognition
- speech processing
- speech recognizer
- pattern recognition
- mel frequency cepstral coefficients
- speech recognition technology
- speaker identification
- cepstral coefficients
- noisy environments
- speech retrieval
- low level
- acoustic features
- acoustic models
- isolated word
- vocal tract
- extracted features
- feature extraction
- visual speech recognition
- image processing
- feature set
- bayesian networks