CTC regularized model adaptation for improving LSTM RNN based multi-accent Mandarin speech recognition.
Jiangyan YiHao NiZhengqi WenBin LiuJianhua TaoPublished in: ISCSLP (2016)
Keyphrases
- speech recognition
- recurrent neural networks
- speaker independent
- language model
- automatic speech recognition
- neural network
- pattern recognition
- speech signal
- speech processing
- hidden markov models
- nearest neighbor
- speech recognizer
- speech recognition technology
- speech synthesis
- speech recognition systems
- feed forward
- noisy environments
- speaker identification
- machine learning
- artificial neural networks
- image processing
- speaker dependent
- word recognition
- speech understanding