石渡祥之佑 / Shonosuke Ishiwatari (Mantra株式会社)
[Webサイト]東京大学情報理工学系研究科修了。博士(情報理工学)。日本学術振興会特別研究員（DC2）、東京大学生産技術研究所特任研究員等を経て、2020年 Mantra株式会社を創業。「世界の言葉で、マンガを届ける。」を合言葉に、マンガ機械翻訳技術の研究開発に従事。
概要
We tackle the problem of machine translation of manga. Manga translation involves two important problems in MT: context-aware and multimodal translation. Since text and images are mixed up in an unstructured fashion in manga, obtaining context from the image is essential for its translation. However, it is still an open problem how to extract context from images and integrate it into MT models. In addition, corpora and benchmarks to train and evaluate such models are currently unavailable. In the talk, we discuss our work that establishes the foundation of manga translation research including i) a multimodal context-aware translation framework, ii) the approach to automatic corpus construction from manga images, and iii) a new benchmark to evaluate manga translation.
※トークは日本語です。