Iterative Shallow Fusion of Backward Language Model for End-to-End Speech Recognition.
Atsunori OgawaTakafumi MoriyaNaoyuki KamoNaohiro TawaraMarc DelcroixPublished in: CoRR (2023)
Keyphrases
- end to end
- speech recognition
- language model
- language modeling
- n gram
- probabilistic model
- document retrieval
- information retrieval
- query expansion
- automatic speech recognition
- mixture model
- question answering
- retrieval model
- speech signal
- test collection
- natural language processing
- noisy environments
- word error rate
- handwriting recognition
- translation model
- image fusion
- speech recognition systems
- query terms
- pattern recognition
- feature vectors
- hand crafted
- image processing