Improving Mandarin End-to-End Speech Recognition With Word N-Gram Language Model.
Jinchuan TianJianwei YuChao WengYuexian ZouDong YuPublished in: IEEE Signal Process. Lett. (2022)
Keyphrases
- speech recognition
- language model
- end to end
- n gram
- language modeling
- language modelling
- speech recognizer
- document retrieval
- language independent
- speech recognition systems
- information retrieval
- speaker independent
- test collection
- probabilistic model
- query expansion
- bag of words
- automatic speech recognition
- speech synthesis
- speech signal
- retrieval model
- variable length
- mixture model
- part of speech
- noisy environments
- word segmentation
- translation model
- word error rate
- out of vocabulary
- query terms
- statistical language modeling
- pseudo relevance feedback
- word level
- relevance model
- cross lingual
- information extraction
- pattern recognition