{"links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00224988","sets":["1164:3027:11083:11173"]},"path":["11173"],"owner":"44499","recid":"224988","title":["顔形状パラメータの逐次更新により個人適応を行う深層学習を用いた視線推定"],"pubdate":{"attribute_name":"公開日","attribute_value":"2023-03-06"},"_buckets":{"deposit":"8ffd45ec-b842-4fd9-834b-3b41f2f6be51"},"_deposit":{"id":"224988","pid":{"type":"depid","value":"224988","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"顔形状パラメータの逐次更新により個人適応を行う深層学習を用いた視線推定","author_link":["594582","594583","594580","594581"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"顔形状パラメータの逐次更新により個人適応を行う深層学習を用いた視線推定"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"視線推定・ユーザ心理","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2023-03-06","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"株式会社国際電気通信基礎技術研究所／立命館大学"},{"subitem_text_value":"株式会社国際電気通信基礎技術研究所"},{"subitem_text_value":"兵庫県立大学"},{"subitem_text_value":"立命館大学"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"ATR / Ritsumeikan University","subitem_text_language":"en"},{"subitem_text_value":"ATR","subitem_text_language":"en"},{"subitem_text_value":"University of Hyogo","subitem_text_language":"en"},{"subitem_text_value":"Ritsumeikan 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/224988/files/IPSJ-HCI23202008.pdf","label":"IPSJ-HCI23202008.pdf"},"date":[{"dateType":"Available","dateValue":"2025-03-06"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HCI23202008.pdf","filesize":[{"value":"3.8 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":"33"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"42d27125-08bf-43c7-9376-c8596f6a7102","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":"施, 真琴"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"内海, 章"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"山添, 大丈"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"李, 周浩"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA1221543X","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":"2188-8760","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"従来のモデルベースの視線推定手法では，顔特徴と眼球の間の関係性のモデル化のために眼球パラメータ情報が付与されたデータが必要とされていた．しかし，眼球パラメータは顔画像を撮影するだけでは取得できず，大規模なデータセットを作成することが難しいという問題があった．一方で深層学習に基づくアピアランスベースの視線推定手法では形状モデルを想定しないため十分な性能が得られない．本論文では我々が提案した逐次更新処理と事前定義計算モジュールを深層学習に基づく視線推定処理に導入することで，この問題を解決する．提案手法では複数の観測を動的に利用することを目的に観測対象に固有の顔形状パラメータを推定に用いて，システムの性能を向上させる．また，顔特徴と眼球・視線方向の間の既知の幾何的な関係性に着目し，正解ラベルとしては視線方向と虹彩中心座標のみを与え，モデルベースの視線推定に必要な眼球パラメータがネットワーク内で誘導されるように学習する．これにより，学習時に眼球パラメータの正解値が不要となる．CG モデルを利用して生成したデータセットを用いた実験によって，顔形状パラメータの逐次更新処理と事前定義計算モジュールの導入による眼球パラメータ（眼球中心・眼球半径）の誘導，および視線推定精度の向上を確認した．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"研究報告ヒューマンコンピュータインタラクション（HCI）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2023-03-06","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"8","bibliographicVolumeNumber":"2023-HCI-202"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:24:30.318403+00:00","updated":"2025-01-19T12:57:55.965188+00:00","id":224988}