{"created":"2025-01-19T01:33:46.422116+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00232737","sets":["1164:4619:11539:11552"]},"path":["11552"],"owner":"44499","recid":"232737","title":["Stable DiffusionのCross Attentionに基づく領域分割へのベクトルノルム導入の検討"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-02-25"},"_buckets":{"deposit":"4fbbcd31-ceb3-4f66-bb8a-af1a36f1a81d"},"_deposit":{"id":"232737","pid":{"type":"depid","value":"232737","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Stable DiffusionのCross Attentionに基づく領域分割へのベクトルノルム導入の検討","author_link":["630756","630754","630755"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Stable DiffusionのCross Attentionに基づく領域分割へのベクトルノルム導入の検討"}]},"item_type_id":"4","publish_date":"2024-02-25","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"大阪公立大学大学院情報学研究科"},{"subitem_text_value":"大阪公立大学大学院情報学研究科"},{"subitem_text_value":"大阪公立大学大学院情報学研究科"}]},"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/232737/files/IPSJ-CVIM24237046.pdf","label":"IPSJ-CVIM24237046.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-CVIM24237046.pdf","filesize":[{"value":"1.3 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"20"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"4e73060b-86f1-4469-968b-5268d607d1cc","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2024 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG."}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"宮田, 優一"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"井上, 勝文"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"吉岡, 理文"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11131797","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":"2188-8701","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"Stable Diffusion では，入力画像の各画素と入力テキストの各単語間の関連度を表す Cross Attention を計算しており，そのAttention Weight を用いた領域分割手法が提案されている．またこの領域分割手法では，入力画像の各画素間の関連度を表す Self Attention を用いた，領域分割結果の洗練も行う．領域分割精度を向上するには，Cross Attention の精度向上と Self Attention による洗練方法の改善が課題となる．本研究では前者の課題に対し，Attention Weight 以外の要素も考慮したベクトルノルムという手法の導入を提案する．また，本研究では後者の課題に対し，複数存在する Attention Weight を合計した領域分割結果を洗練する代わりに個別に洗練する手法と，Self Attention に対するベクトルノルムの導入を提案する．本研究では，提案手法の有効性を複数のデータセットで検証する．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告コンピュータビジョンとイメージメディア（CVIM）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2024-02-25","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"46","bibliographicVolumeNumber":"2024-CVIM-237"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":232737,"updated":"2025-01-19T10:20:17.124075+00:00","links":{}}