{"id":199666,"updated":"2025-01-19T21:37:00.649647+00:00","links":{},"created":"2025-01-19T01:03:32.932122+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00199666","sets":["1164:5159:9712:9932"]},"path":["9932"],"owner":"44499","recid":"199666","title":["音声波形直接生成モデル「ニューラルボコーダ」の比較"],"pubdate":{"attribute_name":"公開日","attribute_value":"2019-10-04"},"_buckets":{"deposit":"e7c80456-69a9-4120-a32b-d2ccdf4f9b63"},"_deposit":{"id":"199666","pid":{"type":"depid","value":"199666","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"音声波形直接生成モデル「ニューラルボコーダ」の比較","author_link":["483955","483954","483953","483956"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"音声波形直接生成モデル「ニューラルボコーダ」の比較"},{"subitem_title":"A comparison of neural vocoders directly synthesizing raw speech waveforms","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"チュートリアル講演","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2019-10-04","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"情報通信研究機構"},{"subitem_text_value":"名古屋大学／情報通信研究機構"},{"subitem_text_value":"情報通信研究機構"},{"subitem_text_value":"情報通信研究機構"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"National Institute of Information and Communications Technology (NICT)","subitem_text_language":"en"},{"subitem_text_value":"Nagoya Uniersity / National Institute of Information and Communications Technology (NICT)","subitem_text_language":"en"},{"subitem_text_value":"National Institute of Information and Communications Technology (NICT)","subitem_text_language":"en"},{"subitem_text_value":"National Institute of Information and Communications Technology (NICT)","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/199666/files/IPSJ-SLP19129003.pdf","label":"IPSJ-SLP19129003.pdf"},"date":[{"dateType":"Available","dateValue":"2021-10-04"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP19129003.pdf","filesize":[{"value":"634.8 kB"}],"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":"c42b0484-9d7f-44b3-879a-b7470f996963","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2019 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"岡本, 拓磨"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"戸田, 智基"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"志賀, 芳則"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"河井, 恒"}],"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":"2016年 までの統計的テキスト音声合成や声質変換では，ニューラルネットワークに基づく音響モデルを用いたとしても，ソースフィルタモデルに基づくボコーダによる音質劣化が肉声感を阻む大きな壁となっていた．2016 年 9 月，WaveNet からはじまるニューラルボコーダの登場により，言語特徴量や音響特徴量からニューラルネットにより音声波形を直接合成できるようになり，Tacotron 2 においては，ついに自然音声と区別のつかない品質の英語テキスト音声合成が実現された．現在，ニューラルボコーダは音声合成における基盤技術となり，様々な方式が提案されている．本チュートリアルでは，WaveNet の登場から最先端のリアルタイムニューラルボコーダまでを紹介し，合成精度，合成速度，モデルサイズ，学習難易度，学習時間等の観点からの比較を行う．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告音声言語情報処理（SLP）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2019-10-04","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicVolumeNumber":"2019-SLP-129"}]},"relation_version_is_last":true,"weko_creator_id":"44499"}}