BERT Meets CTC: New Formulation of End-to-End Speech Recognition with Pre-trained Masked Language Model.
Yosuke HiguchiBrian YanSiddhant AroraTetsuji OgawaTetsunori KobayashiShinji WatanabePublished in: CoRR (2022)
Keyphrases
- end to end
- speech recognition
- language model
- pre trained
- language modeling
- probabilistic model
- n gram
- document retrieval
- training data
- information retrieval
- automatic speech recognition
- word error rate
- training examples
- test collection
- query expansion
- retrieval model
- speech signal
- mixture model
- query terms
- translation model
- computer vision
- hidden markov models
- decision trees
- handwriting recognition
- machine learning