Conversion of Recurrent Neural Network Language Models to Weighted Finite State Transducers for Automatic Speech Recognition.
Gwénolé LecorvéPetr MotlícekPublished in: INTERSPEECH (2012)
Keyphrases
- language model
- recurrent neural networks
- automatic speech recognition
- speech recognition
- language modeling
- finite state transducers
- word error rate
- n gram
- speech signal
- feed forward
- retrieval model
- neural network
- document retrieval
- probabilistic model
- test collection
- query expansion
- broadcast news
- information retrieval
- artificial neural networks
- language modelling
- query terms
- hidden markov models
- cross lingual
- translation model
- word recognition
- bayesian networks
- data mining
- relevance model
- pseudo relevance feedback
- pattern recognition
- back propagation
- computer vision
- word segmentation
- language independent
- term dependencies
- information extraction
- natural language processing