Incoherent training of deep neural networks to de-correlate bottleneck features for speech recognition.
Yebo BaoHui JiangLi-Rong DaiCong LiuPublished in: ICASSP (2013)
Keyphrases
- speech recognition
- neural network
- pattern recognition
- speech recognition systems
- hidden markov models
- wall street journal corpus
- speech processing
- isolated word
- speech synthesis
- speech signal
- automatic speech recognition
- feature extraction
- training process
- speech recognizer
- speech understanding
- noisy environments
- language model
- training set
- cepstral coefficients
- speech recognition technology
- speech retrieval
- speaker diarization
- speaker independent
- speaker dependent
- feature space