Minimum Classification Error for Large Scale Speech Recognition Tasks using Weighted Finite State Transducers.
Erik McDermottShigeru KatagiriPublished in: ICASSP (1) (2005)
Keyphrases
- finite state transducers
- minimum classification error
- statistical pattern recognition
- finite state
- machine translation
- language modeling
- n gram
- hidden markov models
- grammatical inference
- speech recognition
- edit distance
- language model
- discriminative training
- neural network
- distance function
- speech signal
- query expansion
- denoising