{"created":"2025-01-19T01:39:24.340215+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00237016","sets":["6504:11678:11684"]},"path":["11684"],"owner":"44499","recid":"237016","title":["日本近代公文書画像における文脈を考慮した文字検出手法"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-03-01"},"_buckets":{"deposit":"bbba6adf-fc9e-493c-91ec-71be6b1210d4"},"_deposit":{"id":"237016","pid":{"type":"depid","value":"237016","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"日本近代公文書画像における文脈を考慮した文字検出手法","author_link":["648093","648097","648096","648094","648095","648092"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"日本近代公文書画像における文脈を考慮した文字検出手法"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"コンピュータと人間社会","subitem_subject_scheme":"Other"}]},"item_type_id":"22","publish_date":"2024-03-01","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_22_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"中京大"},{"subitem_text_value":"中京大"},{"subitem_text_value":"中京大"},{"subitem_text_value":"中京大"},{"subitem_text_value":"中京大"},{"subitem_text_value":"中京大"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/237016/files/IPSJ-Z86-7ZK-05.pdf","label":"IPSJ-Z86-7ZK-05.pdf"},"date":[{"dateType":"Available","dateValue":"2024-07-04"}],"format":"application/pdf","filename":"IPSJ-Z86-7ZK-05.pdf","filesize":[{"value":"499.6 kB"}],"mimetype":"application/pdf","accessrole":"open_date","version_id":"8fee2baa-14a3-4918-9fc2-8e5078a54375","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2024 by the Information Processing Society of Japan"}]},"item_22_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"宮川, 裕貴"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"山田, 雅之"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"中, 貴俊"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"兼松, 篤子"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"宮崎, 慎也"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"長谷川, 純一"}],"nameIdentifiers":[{}]}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_5794","resourcetype":"conference paper"}]},"item_22_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN00349328","subitem_source_identifier_type":"NCID"}]},"item_22_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近代の公文書には史料的価値があるが，その多くはくずし字を含む手書きの文書であるため近代古文書の知識が無ければ解読が困難である．我々は台湾総督府文書を題材として，近代公文書を対象としたOCRシステムの開発を進めている．本研究では文字検出を行う機械学習モデルとして，画像特徴と文脈情報を利用するTransformerベースのモデルを提案する．このモデルはエンコーダでは画像特徴の抽出を行い，デコーダでは画像特徴と文脈情報を合わせて文字検出のための特徴を獲得，これを利用して画像中の各文字の領域と字種を予測する．近代公文書画像に対して文脈を考慮しないモデルとの精度比較を通じて提案モデルの有効性を示す．","subitem_description_type":"Other"}]},"item_22_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"818","bibliographic_titles":[{"bibliographic_title":"第86回全国大会講演論文集"}],"bibliographicPageStart":"817","bibliographicIssueDates":{"bibliographicIssueDate":"2024-03-01","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"1","bibliographicVolumeNumber":"2024"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":237016,"updated":"2025-01-19T09:01:53.990866+00:00","links":{}}