{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00213080","sets":["6164:6165:6640:10712"]},"path":["10712"],"owner":"44499","recid":"213080","title":["動作認識のための合成データ活用に向けたドメイン適応手法の比較"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-06-23"},"_buckets":{"deposit":"0247222e-0c07-445b-828b-33a9cfb98c60"},"_deposit":{"id":"213080","pid":{"type":"depid","value":"213080","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"動作認識のための合成データ活用に向けたドメイン適応手法の比較","author_link":["544591","544590","544592","544593"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"動作認識のための合成データ活用に向けたドメイン適応手法の比較"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"AI","subitem_subject_scheme":"Other"}]},"item_type_id":"18","publish_date":"2021-06-23","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_18_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"お茶の水女子大学"},{"subitem_text_value":"国立情報学研究所"},{"subitem_text_value":"産業技術総合研究所"},{"subitem_text_value":"お茶の水女子大学"}]},"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/213080/files/IPSJ-DICOMO2021182.pdf","label":"IPSJ-DICOMO2021182.pdf"},"date":[{"dateType":"Available","dateValue":"2023-06-23"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-DICOMO2021182.pdf","filesize":[{"value":"2.6 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"30acec5f-32c9-4496-9bf7-ff08e3fb573d","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_18_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"礒井, 葉那"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"竹房, あつ子"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"中田, 秀基"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"小口, 正人"}],"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_18_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"ディープニューラルネットワークの進歩に伴う学習データ不足の問題について様々な議論が行われており，その解決策の 1 つに合成データを利用した学習がある．合成データには生成が比較的容易であるという利点があるが，合成データを用いて学習したモデルには，実データ解析時にドメインシフトによって解析精度が低下するという課題がある．本研究では，合成動画像データを活用した高精度な実動画像データ識別の実現を目的とし，写実的な合成動画像データを用いて 3D ResNet と TSN をベースとするモデルでそれぞれ学習し，その動作識別精度を比較した. 実験の結果，合成データと実データの特徴の違いはモーションよりも色や形状，質感にあること，オプティカルフローを用いる TSN ベースのモデルの方が高精度に実データの動作識別が可能であることがわかった．","subitem_description_type":"Other"}]},"item_18_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"1297","bibliographic_titles":[{"bibliographic_title":"マルチメディア，分散協調とモバイルシンポジウム2021論文集"}],"bibliographicPageStart":"1289","bibliographicIssueDates":{"bibliographicIssueDate":"2021-06-23","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"1","bibliographicVolumeNumber":"2021"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":213080,"updated":"2025-01-19T17:16:58.183898+00:00","links":{},"created":"2025-01-19T01:13:59.665793+00:00"}