Recurrent neural network language model training with noise contrastive estimation for speech recognition.
Xie ChenXunying LiuMark J. F. GalesPhilip C. WoodlandPublished in: ICASSP (2015)
Keyphrases
- speech recognition
- language model
- recurrent neural networks
- recurrent networks
- noisy environments
- language modeling
- isolated word
- n gram
- acoustic models
- document retrieval
- speech recognizer
- probabilistic model
- retrieval model
- feed forward
- information retrieval
- neural network
- automatic speech recognition
- speech signal
- mixture model
- artificial neural networks
- query expansion
- word error rate
- hidden markov models
- test collection
- training process
- pattern recognition
- speech recognition systems
- signal to noise ratio
- handwriting recognition
- artificial intelligence
- translation model
- relevance model
- error rate
- machine learning