Improving Automatic Evaluation of Mandarin Pronunciation with Speaker Adaptive Training (SAT) and MLLR Speaker Adaption.
Chao HuangFeng ZhangFrank K. SoongPublished in: ISCSLP (2008)
Keyphrases
- speech recognition
- automatic evaluation
- speaker adaptation
- acoustic models
- speaker independent
- automatic speech recognition
- hidden markov models
- speaker identification
- broadcast news
- prosodic features
- speaker verification
- speech recognizer
- noisy environments
- maximum likelihood
- speech synthesis
- speech signal
- quality assessment
- speech recognition systems
- spontaneous speech
- supervised learning
- information retrieval