LLaST: Improved End-to-end Speech Translation System Leveraged by Large Language Models.
Xi ChenSongyang ZhangQibing BaiKai ChenSatoshi NakamuraPublished in: CoRR (2024)
Keyphrases
- end to end
- language model
- speech recognition
- translation model
- word error rate
- language modeling
- cross language retrieval
- out of vocabulary
- probabilistic model
- document retrieval
- n gram
- retrieval model
- query expansion
- spoken term detection
- information retrieval
- statistical language models
- context sensitive
- automatic speech recognition
- congestion control
- smoothing methods
- speech signal
- cross language information retrieval
- pseudo relevance feedback
- language modelling
- admission control
- query terms
- vector space model
- okapi bm
- test collection
- document ranking
- language models for information retrieval
- machine translation
- error rate
- text retrieval
- query translation
- cross language