This paper describes NAIST’s NMT system submitted to the IWSLT 2020 conversational speech translation task. We focus on the translation disfluent speech transcripts that include ASR errors and non-grammatical utterances. We tried a domain adaptation method by transferring the styles of out-of-domain data (United Nations Parallel Corpus) to be like in-domain data (Fisher transcripts). Our system results showed that the NMT model with domain adaptation outperformed a baseline. In addition, slight improvement by the style transfer was observed.
@inproceedings{fukuda-etal-2020-naists, title = "{NAIST}`s Machine Translation Systems for {IWSLT} 2020 Conversational Speech Translation Task", author = "Fukuda, Ryo and Sudoh, Katsuhito and Nakamura, Satoshi", editor = {Federico, Marcello and Waibel, Alex and Knight, Kevin and Nakamura, Satoshi and Ney, Hermann and Niehues, Jan and St{\"u}ker, Sebastian and Wu, Dekai and Mariani, Joseph and Yvon, Francois}, booktitle = "Proceedings of the 17th International Conference on Spoken Language Translation", month = jul, year = "2020", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2020.iwslt-1.21/", doi = "10.18653/v1/2020.iwslt-1.21", pages = "172--177", abstract = "This paper describes NAIST`s NMT system submitted to the IWSLT 2020 conversational speech translation task. We focus on the translation disfluent speech transcripts that include ASR errors and non-grammatical utterances. We tried a domain adaptation method by transferring the styles of out-of-domain data (United Nations Parallel Corpus) to be like in-domain data (Fisher transcripts). Our system results showed that the NMT model with domain adaptation outperformed a baseline. In addition, slight improvement by the style transfer was observed."}
%0 Conference Proceedings%T NAIST‘s Machine Translation Systems for IWSLT 2020 Conversational Speech Translation Task%A Fukuda, Ryo%A Sudoh, Katsuhito%A Nakamura, Satoshi%Y Federico, Marcello%Y Waibel, Alex%Y Knight, Kevin%Y Nakamura, Satoshi%Y Ney, Hermann%Y Niehues, Jan%Y Stüker, Sebastian%Y Wu, Dekai%Y Mariani, Joseph%Y Yvon, Francois%S Proceedings of the 17th International Conference on Spoken Language Translation%D 2020%8 July%I Association for Computational Linguistics%C Online%F fukuda-etal-2020-naists%X This paper describes NAIST‘s NMT system submitted to the IWSLT 2020 conversational speech translation task. We focus on the translation disfluent speech transcripts that include ASR errors and non-grammatical utterances. We tried a domain adaptation method by transferring the styles of out-of-domain data (United Nations Parallel Corpus) to be like in-domain data (Fisher transcripts). Our system results showed that the NMT model with domain adaptation outperformed a baseline. In addition, slight improvement by the style transfer was observed.%R 10.18653/v1/2020.iwslt-1.21%U https://aclanthology.org/2020.iwslt-1.21/%U https://doi.org/10.18653/v1/2020.iwslt-1.21%P 172-177