and interpret business documents, collectively referred to as “Document Intelligence” ϏδωεจॻΛಡΈɺཧղ͠ɺղऍ͢ΔೳྗΛɺ૯শͯ͠ʮυΩϡϝϯτɾ ΠϯςϦδΣϯεʯͱݺͼ·͢ “Workshop on Document Intelligence”ΑΓ https://sites.google.com/view/di2019/home
• ҰํͰɺݸʑͷٕज़͕σΟʔϓϥʔχϯάʹΑΓਫ਼্͍ͯ͠Δ • จࣈಡΈऔΓਫ਼Λ্ͤ͞ΔΑ͏ͳը૾ͷิਖ਼ٕज़ʢղ૾ʣ • ຊޠͷݴޠϞσϧతʹલޙͷจ຺Λߟྀͨ͠OCR • จࣈྻʹՃ͑ͯը૾্ͷ࠲ඪใΛಉ࣌ʹΈࠐΜͩॻྨ൛BERT (Layout LM) ݸʑͷٕज़͕·ͩ·ͩൃల్্ *1 https://speakerdeck.com/sansandsoc/recent-topics-on-character-super-resolution *2 https://speakerdeck.com/line_devday2019/naver-clova-ocr *3 Xu, Yiheng, et al. "Layoutlm: Pre-training of text and layout for document image understanding." Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2020.