探討聲學模型的合併技術與半監督鑑別式訓練於會議語音辨識之研究 (Investigating acoustic model combination and semi-supervised discriminative training for meeting speech recognition) [In Chinese].
Tien-Hong LoBerlin ChenPublished in: ROCLING (2018)
Keyphrases
- speech recognition
- discriminative training
- automatic speech recognition
- hidden markov models
- semi supervised
- co training
- speech synthesis
- acoustic models
- speaker diarization
- speech signal
- language model
- semi supervised learning
- labeled data
- pattern recognition
- noisy environments
- unlabeled data
- multi view
- speech recognition systems
- probabilistic interpretation
- speech recognizer
- word recognition
- broadcast news
- speaker identification
- pairwise
- supervised learning
- speaker independent
- active learning
- maximum likelihood
- probabilistic model
- n gram
- neural network