Improved tone modeling by exploiting articulatory features for mandarin speech recognition.
Hao ChaoZhanlei YangWenju LiuPublished in: ICASSP (2012)
Keyphrases
- speech recognition
- speech recognition systems
- hidden markov models
- speech synthesis
- speech signal
- automatic speech recognition
- language model
- speech recognizer
- pattern recognition
- noisy environments
- speaker identification
- speech processing
- cepstral coefficients
- speaker independent
- speech recognition technology
- speech understanding
- feature extraction
- speaker dependent
- feature set
- feature vectors
- speech retrieval
- low level
- keyword spotting
- bayesian networks
- speech recognition errors
- feature selection