FST: the FAIR Speech Translation System for the IWSLT21 Multilingual Shared Task.
Yun TangHongyu GongXian LiChanghan WangJuan Miguel PinoHolger SchwenkNaman GoyalPublished in: IWSLT (2021)
Keyphrases
- language resources
- cross language information retrieval
- broadcast news
- multi lingual
- cross language
- cross language ir
- machine translation
- cross lingual information retrieval
- speech recognition
- machine translation system
- cross lingual
- parallel corpus
- automatic speech recognition
- comparable corpora
- query translation
- chinese english
- language independent
- parallel corpora
- finite state transducers
- speech signal
- recognition engine
- test set
- digital libraries
- bilingual dictionaries
- text to speech
- speech synthesis
- out of vocabulary
- document collections
- word alignment
- named entity recognition
- audio visual
- text retrieval
- speaker identification
- document retrieval
- translation model
- coreference resolution
- information access
- syntactic and semantic dependencies