{"links":{},"id":57424,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00057424","sets":["1164:5159:5204:5208"]},"path":["5208"],"owner":"1","recid":"57424","title":["話し言葉音声認識のための言語モデルとデコーダの改善"],"pubdate":{"attribute_name":"公開日","attribute_value":"2001-06-01"},"_buckets":{"deposit":"14de2cbf-271d-45ae-b491-657b3a7406ae"},"_deposit":{"id":"57424","pid":{"type":"depid","value":"57424","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"話し言葉音声認識のための言語モデルとデコーダの改善","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"話し言葉音声認識のための言語モデルとデコーダの改善"},{"subitem_title":"Language Model and Decoder for Lecture Speech Recogntion","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2001-06-01","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"京都大学情報学研究科"},{"subitem_text_value":"京都大学情報学研究科"},{"subitem_text_value":"京都大学情報学研究科"},{"subitem_text_value":"奈良先端科学技術大学院大学情報科学研究科"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"Nara Institute of Science and Technology","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/57424/files/IPSJ-SLP01036003.pdf"},"date":[{"dateType":"Available","dateValue":"2003-06-01"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP01036003.pdf","filesize":[{"value":"150.2 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":"26dcc095-7d84-463a-bd83-c768361e9013","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2001 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"河原, 達也"},{"creatorName":"加藤, 一臣"},{"creatorName":"南條浩輝"},{"creatorName":"李晃伸"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Tatsuya, Kawahara","creatorNameLang":"en"},{"creatorName":"Kazuomi, Kato","creatorNameLang":"en"},{"creatorName":"Hiroaki, Nanjo","creatorNameLang":"en"},{"creatorName":"Akinobu, Lee","creatorNameLang":"en"}],"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_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"講演音声のような話し言葉を認識するための言語モデルとデコーダ(音声認識プログラム)の改善法について述べる。言語モデルについては、話し言葉コーパスのデータ量不足を補うために、他のコーパスと混合する方法、特に混合重みの最適化手法を考察する。また話し言葉では、ポーズで区切られる単位が言語的な単位と一致しないため、間投詞とあわせて扱いの検討が必要である。特に発話のセグメンテーションが容易でないため、ショートポーズを検出しながら逐次確定していく方式をデコーダJuliusに実装した。10名の講演音声の認識で評価を行い、平均65.3%の認識精度を得た。","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Language modeling and decoding algorithms for spontaneous lecture speech recognition are addressed.  In order to complement training data for the language model, incorporation of other text corpora is explored.  Optimization of the text mixture weights is proposed. Handling of pauses as well as fillers is another problem in spontaneous speech recognition.  We revised our decoder {\\sf Julius} so that speech segmentation is performed along the recognition process.  We have achieved word accuracy of 65.3\\% on the automatic transcription of real oral presentations.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"22","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告音声言語情報処理（SLP）"}],"bibliographicPageStart":"15","bibliographicIssueDates":{"bibliographicIssueDate":"2001-06-01","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"55(2001-SLP-036)","bibliographicVolumeNumber":"2001"}]},"relation_version_is_last":true,"weko_creator_id":"1"},"created":"2025-01-18T23:20:37.836483+00:00","updated":"2025-01-22T04:30:21.000522+00:00"}