{"links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00228842","sets":["1164:2036:11089:11357"]},"path":["11357"],"owner":"44499","recid":"228842","title":["自動運転に向けた車載カメラ画像からの物体検出タスクへの連合学習の活用"],"pubdate":{"attribute_name":"公開日","attribute_value":"2023-10-31"},"_buckets":{"deposit":"353b7a94-4582-41b1-9d6d-c0a57ce50597"},"_deposit":{"id":"228842","pid":{"type":"depid","value":"228842","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"自動運転に向けた車載カメラ画像からの物体検出タスクへの連合学習の活用","author_link":["614290","614292","614289","614291"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"自動運転に向けた車載カメラ画像からの物体検出タスクへの連合学習の活用"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"ポスター","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2023-10-31","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"大阪大学大学院情報科学研究科"},{"subitem_text_value":"大阪大学大学院情報科学研究科"},{"subitem_text_value":"大阪大学大学院情報科学研究科"},{"subitem_text_value":"大阪大学大学院情報科学研究科"}]},"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/228842/files/IPSJ-SLDM23203007.pdf","label":"IPSJ-SLDM23203007.pdf"},"date":[{"dateType":"Available","dateValue":"2025-10-31"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLDM23203007.pdf","filesize":[{"value":"2.5 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":"10"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"23e0cc66-a99a-474c-8411-a275f5000319","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2023 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"澤, 風馬"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"西川, 広記"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"谷口, 一徹"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"尾上, 孝雄"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11451459","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8639","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"自動運転では，車載カメラなどからの画像を利用して人検出や車両検出などが行われている．他方，それに向けた物体検出モデルを構築する上で様々な車両からの画像が求められるが，一般ユーザからの車載カメラ画像の収集はプライバシ漏洩につながる可能性が指摘されており，近年ではプライバシ保護を考慮した分散型機械学習のひとつとして，連合学習が注目されている．しかしながら，複数のクライアントに分散してモデルを学習する性質上，連合学習は学習データの非均質性が一般に課題とされる．そこで本稿では，車載カメラ画像からの物体検出の精度を向上させるための連合学習手法を提案する．提案手法における初期検討として，本稿では各クライアントにおける学習用の車載カメラ画像がなるべく均質になるよう昼画像クラスタと夜画像クラスタとに分解された状況を想定し，画像クラスタ毎に連合学習を行う．実験では，提案手法とクラス分けを行わない従来手法とを比較し，状況に応じてモデルを使い分けることの重要性が示された．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"5","bibliographic_titles":[{"bibliographic_title":"研究報告システムとLSIの設計技術（SLDM）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2023-10-31","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"7","bibliographicVolumeNumber":"2023-SLDM-203"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:27:58.636439+00:00","updated":"2025-01-19T11:41:57.253201+00:00","id":228842}