{"updated":"2025-01-19T09:42:47.263095+00:00","links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00234724","sets":["1164:5159:11541:11627"]},"path":["11627"],"owner":"44499","recid":"234724","title":["Beyond Word Count: Exploring Approximated Target Lengths for CIF-RNNT"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-06-07"},"_buckets":{"deposit":"8955f4a1-616c-449a-baff-5d38d5264c4f"},"_deposit":{"id":"234724","pid":{"type":"depid","value":"234724","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Beyond Word Count: Exploring Approximated Target Lengths for CIF-RNNT","author_link":["639999","640002","640001","640000"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Beyond Word Count: Exploring Approximated Target Lengths for CIF-RNNT"},{"subitem_title":"Beyond Word Count: Exploring Approximated Target Lengths for CIF-RNNT","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"ポスターセッション1","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2024-06-07","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"The University of Electro-Communications"},{"subitem_text_value":"The University of Electro-Communications"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"The University of Electro-Communications","subitem_text_language":"en"},{"subitem_text_value":"The University of Electro-Communications","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"eng"}]},"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/234724/files/IPSJ-SLP24152037.pdf","label":"IPSJ-SLP24152037.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP24152037.pdf","filesize":[{"value":"1.9 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"22"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"e480596e-6a09-445a-93c5-b5fa24b601a8","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":"Wen, Shen Teo"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Yasuhiro, Minami"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Wen, Shen Teo","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Yasuhiro, Minami","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_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":"Our previous work proposed the CIF-RNNT architecture, a combination of Continuous Integrate-and-Fire (CIF) and RNN-Transducers (RNN-T) that compresses speech into units equivalent to linguistic words to achieve efficient decoding. This work extends on that research by investigating the impact of different target length definitions, approximated from self-information and token count. Our results on LibriSpeech and CSJ show that approximated target length types based on self-information outperform simpler approaches, and CIF-RNNT models even surpass topline models on the CSJ dataset at smaller chunk sizes. Furthermore, our comparisons demonstrate an inherent ability of CIF-RNNT to produce output tokens in groups of words, regardless of the target length type. These results showcase the potential of the CIF-RNNT architecture for efficient and accurate speech recognition.","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Our previous work proposed the CIF-RNNT architecture, a combination of Continuous Integrate-and-Fire (CIF) and RNN-Transducers (RNN-T) that compresses speech into units equivalent to linguistic words to achieve efficient decoding. This work extends on that research by investigating the impact of different target length definitions, approximated from self-information and token count. Our results on LibriSpeech and CSJ show that approximated target length types based on self-information outperform simpler approaches, and CIF-RNNT models even surpass topline models on the CSJ dataset at smaller chunk sizes. Furthermore, our comparisons demonstrate an inherent ability of CIF-RNNT to produce output tokens in groups of words, regardless of the target length type. These results showcase the potential of the CIF-RNNT architecture for efficient and accurate speech recognition.","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":"2024-06-07","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"37","bibliographicVolumeNumber":"2024-SLP-152"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":234724,"created":"2025-01-19T01:36:36.857086+00:00"}