{"links":{},"id":231310,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00231310","sets":["1164:5159:11151:11431"]},"path":["11431"],"owner":"44499","recid":"231310","title":["アダプタを用いた大規模事前学習モデルの話者適応"],"pubdate":{"attribute_name":"公開日","attribute_value":"2023-11-25"},"_buckets":{"deposit":"c338ec97-6435-4c2d-91f7-6dd0ec6f9ce8"},"_deposit":{"id":"231310","pid":{"type":"depid","value":"231310","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"アダプタを用いた大規模事前学習モデルの話者適応","author_link":["624257","624256"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"アダプタを用いた大規模事前学習モデルの話者適応"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"分野横断（2）","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2023-11-25","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"京都大学大学院情報学研究科"},{"subitem_text_value":"京都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/231310/files/IPSJ-SLP23149018.pdf","label":"IPSJ-SLP23149018.pdf"},"date":[{"dateType":"Available","dateValue":"2025-11-25"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP23149018.pdf","filesize":[{"value":"1.4 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":"22"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"823a7798-bd77-48df-8a08-99f97dd0ea51","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":"渋谷, 和樹"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"河原, 達也"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10442647","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":"2188-8663","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"音声認識を用いたアプリケーションの普及に伴い，各話者に適応した音声認識モデルの需要が高まっている．しかし，深層学習を用いた音声認識モデルは一般にパラメータ数が多く，各話者専用のモデルを保持することはメモリの観点から難しい．そこで，本研究では少ないパラメータ数で個別の話者に適応した音声認識モデルを提供する方法として，アダプタと呼ばれる小規模な追加ネットワークを用いた話者適応を実装する．また，実応用において，各話者の発話を大量に収集することは難しい．そこで，正解ラベルを用いてアダプタを学習する教師あり学習のみでなく，事前学習済みモデルの推論結果を利用する自己ラベル学習，および書き起こしを利用しない自己教師あり学習についても検討する．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"5","bibliographic_titles":[{"bibliographic_title":"研究報告音声言語情報処理（SLP）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2023-11-25","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"18","bibliographicVolumeNumber":"2023-SLP-149"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:31:32.204005+00:00","updated":"2025-01-19T10:48:54.948686+00:00"}