{"id":217545,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00217545","sets":["934:989:10777:10860"]},"path":["10860"],"owner":"44499","recid":"217545","title":["弱教師あり学習による連続的な表情特徴の獲得"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-03-16"},"_buckets":{"deposit":"bc5a0f75-a76d-435d-bc6a-1c1c7004ec21"},"_deposit":{"id":"217545","pid":{"type":"depid","value":"217545","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"弱教師あり学習による連続的な表情特徴の獲得","author_link":["563823","563825","563824","563822"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"弱教師あり学習による連続的な表情特徴の獲得"},{"subitem_title":"Weakly Supervised Learning for Acquisition of Continuous Facial Expression Features","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"[オリジナル論文] 弱教師あり学習，縺れを解いた特徴表現学習，表情認識，画像生成，変分オートエンコーダ，敵対的生成ネットワーク","subitem_subject_scheme":"Other"}]},"item_type_id":"3","publish_date":"2022-03-16","item_3_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"横浜国立大学大学院環境情報学府"},{"subitem_text_value":"横浜国立大学大学院環境情報研究院"}]},"item_3_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of Environment and Information Sciences, Yokohama National University","subitem_text_language":"en"},{"subitem_text_value":"Faculty of Environment and Information Sciences, Yokohama National University","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"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/217545/files/IPSJ-TOM1502003.pdf","label":"IPSJ-TOM1502003.pdf"},"date":[{"dateType":"Available","dateValue":"2024-03-16"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-TOM1502003.pdf","filesize":[{"value":"1.6 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":"17"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"a492aa11-51d5-4201-b167-bc45cf654a73","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 by the Information Processing Society of Japan"}]},"item_3_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"狩野, 悌久"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"長尾, 智晴"}],"nameIdentifiers":[{}]}]},"item_3_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yoshihisa, Kanou","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Tomoharu, Nagao","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_3_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11464803","subitem_source_identifier_type":"NCID"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_6501","resourcetype":"journal article"}]},"item_3_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7780","subitem_source_identifier_type":"ISSN"}]},"item_3_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"機械学習による表情特徴の獲得は，一般的に感情ラベルなどを用いた教師あり学習によって行われる．しかし，人間による表情のラベリングは，主観的なものになりやすく教師自体が曖昧性を持つ可能性がある．また，表情に対して感情のクラスを割り当てることは連続的な表情を離散的に扱うことになり，モデルが表情の連続性を学習することを妨げる．そこで私たちは，表情に関連する教師情報を用いることなく，被験者特徴から切り離された状態の表情特徴を獲得することを研究の目的とした．本稿では，以前私たちが提案した手法に対して2種類の損失関数を導入し，さらに学習プロセスを改良することにより，しわなどの細かな情報を含んだ表情特徴を獲得する手法を提案する．実験では，表情認識と画像生成を行い，提案手法によって獲得された特徴量の有効性を示す．","subitem_description_type":"Other"}]},"item_3_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Acquisition of facial expression features from images by machine learning is generally done with supervision, such as using emotional labels that match facial expressions. However, in the supervised setting, there are problems such as ambiguity of the label due to the subjectivity of the person to the facial expression and the discrete treatment of continuous facial expressions by giving the label. To solve these problems, we focus on acquiring continuous facial expression features without using information of facial expression for training the model. In this paper, we improve the weakly supervised method proposed in “Separation of the Latent Representations into Identity and Expression without Emotional Labels” to acquire more effective facial expression features. The experimental result shows that the proposed method acquire effective facial expression features and achieve better results than the previous method in each task of image generation and facial expression recognition.","subitem_description_type":"Other"}]},"item_3_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"20","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌数理モデル化と応用（TOM）"}],"bibliographicPageStart":"11","bibliographicIssueDates":{"bibliographicIssueDate":"2022-03-16","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"2","bibliographicVolumeNumber":"15"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"updated":"2025-01-19T15:27:32.288968+00:00","created":"2025-01-19T01:18:02.125222+00:00","links":{}}