Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus

The punctuation restoration task aims to correctly punctuate the output transcriptions of automatic speech recognition systems. Previous punctuation models, either using text only or demanding the corresponding audio, tend to be constrained by real scenes, where unpunctuated sentences are a mixture...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhu, Yaoming, Wu, Liwei, Cheng, Shanbo, Wang, Mingxuan
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Zhu, Yaoming
Wu, Liwei
Cheng, Shanbo
Wang, Mingxuan
description The punctuation restoration task aims to correctly punctuate the output transcriptions of automatic speech recognition systems. Previous punctuation models, either using text only or demanding the corresponding audio, tend to be constrained by real scenes, where unpunctuated sentences are a mixture of those with and without audio. This paper proposes a unified multimodal punctuation restoration framework, named UniPunc, to punctuate the mixed sentences with a single model. UniPunc jointly represents audio and non-audio samples in a shared latent space, based on which the model learns a hybrid representation and punctuates both kinds of samples. We validate the effectiveness of the UniPunc on real-world datasets, which outperforms various strong baselines (e.g. BERT, MuSe) by at least 0.8 overall F1 scores, making a new state-of-the-art. Extensive experiments show that UniPunc's design is a pervasive solution: by grafting onto previous models, UniPunc enables them to punctuate on the mixed corpus. Our code is available at github.com/Yaoming95/UniPunc
doi_str_mv 10.48550/arxiv.2202.00468
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2202_00468</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2202_00468</sourcerecordid><originalsourceid>FETCH-LOGICAL-a678-e8b7877a2698d39d171fe4e31f2e67deb4b5b33153aa3bd5f92315d8eb4182873</originalsourceid><addsrcrecordid>eNotj8tOwzAURL1hgQofwAr_QIIfcewsUUQBqSkItevour6WLJK4chJo_54-WM2MRjPSIeSBs7wwSrEnSIfwkwvBRM5YUZpbst4OwQd0tJm7KfTRQUc_52E3zTCFONAvHKeYrn6ZoMffmL6pj4k24YAua86LMB1pHdN-Hu_IjYduxPt_XZDN8mVTv2Wrj9f3-nmVQalNhsZqozWIsjJOVo5r7rFAyb3AUju0hVVWSq4kgLRO-UqcgjOnghthtFyQx-vtBajdp9BDOrZnsPYCJv8AIrhJdg</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus</title><source>arXiv.org</source><creator>Zhu, Yaoming ; Wu, Liwei ; Cheng, Shanbo ; Wang, Mingxuan</creator><creatorcontrib>Zhu, Yaoming ; Wu, Liwei ; Cheng, Shanbo ; Wang, Mingxuan</creatorcontrib><description>The punctuation restoration task aims to correctly punctuate the output transcriptions of automatic speech recognition systems. Previous punctuation models, either using text only or demanding the corresponding audio, tend to be constrained by real scenes, where unpunctuated sentences are a mixture of those with and without audio. This paper proposes a unified multimodal punctuation restoration framework, named UniPunc, to punctuate the mixed sentences with a single model. UniPunc jointly represents audio and non-audio samples in a shared latent space, based on which the model learns a hybrid representation and punctuates both kinds of samples. We validate the effectiveness of the UniPunc on real-world datasets, which outperforms various strong baselines (e.g. BERT, MuSe) by at least 0.8 overall F1 scores, making a new state-of-the-art. Extensive experiments show that UniPunc's design is a pervasive solution: by grafting onto previous models, UniPunc enables them to punctuate on the mixed corpus. Our code is available at github.com/Yaoming95/UniPunc</description><identifier>DOI: 10.48550/arxiv.2202.00468</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computation and Language</subject><creationdate>2022-01</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,777,882</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2202.00468$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2202.00468$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Zhu, Yaoming</creatorcontrib><creatorcontrib>Wu, Liwei</creatorcontrib><creatorcontrib>Cheng, Shanbo</creatorcontrib><creatorcontrib>Wang, Mingxuan</creatorcontrib><title>Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus</title><description>The punctuation restoration task aims to correctly punctuate the output transcriptions of automatic speech recognition systems. Previous punctuation models, either using text only or demanding the corresponding audio, tend to be constrained by real scenes, where unpunctuated sentences are a mixture of those with and without audio. This paper proposes a unified multimodal punctuation restoration framework, named UniPunc, to punctuate the mixed sentences with a single model. UniPunc jointly represents audio and non-audio samples in a shared latent space, based on which the model learns a hybrid representation and punctuates both kinds of samples. We validate the effectiveness of the UniPunc on real-world datasets, which outperforms various strong baselines (e.g. BERT, MuSe) by at least 0.8 overall F1 scores, making a new state-of-the-art. Extensive experiments show that UniPunc's design is a pervasive solution: by grafting onto previous models, UniPunc enables them to punctuate on the mixed corpus. Our code is available at github.com/Yaoming95/UniPunc</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computation and Language</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj8tOwzAURL1hgQofwAr_QIIfcewsUUQBqSkItevour6WLJK4chJo_54-WM2MRjPSIeSBs7wwSrEnSIfwkwvBRM5YUZpbst4OwQd0tJm7KfTRQUc_52E3zTCFONAvHKeYrn6ZoMffmL6pj4k24YAua86LMB1pHdN-Hu_IjYduxPt_XZDN8mVTv2Wrj9f3-nmVQalNhsZqozWIsjJOVo5r7rFAyb3AUju0hVVWSq4kgLRO-UqcgjOnghthtFyQx-vtBajdp9BDOrZnsPYCJv8AIrhJdg</recordid><startdate>20220124</startdate><enddate>20220124</enddate><creator>Zhu, Yaoming</creator><creator>Wu, Liwei</creator><creator>Cheng, Shanbo</creator><creator>Wang, Mingxuan</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20220124</creationdate><title>Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus</title><author>Zhu, Yaoming ; Wu, Liwei ; Cheng, Shanbo ; Wang, Mingxuan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a678-e8b7877a2698d39d171fe4e31f2e67deb4b5b33153aa3bd5f92315d8eb4182873</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computation and Language</topic><toplevel>online_resources</toplevel><creatorcontrib>Zhu, Yaoming</creatorcontrib><creatorcontrib>Wu, Liwei</creatorcontrib><creatorcontrib>Cheng, Shanbo</creatorcontrib><creatorcontrib>Wang, Mingxuan</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Zhu, Yaoming</au><au>Wu, Liwei</au><au>Cheng, Shanbo</au><au>Wang, Mingxuan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus</atitle><date>2022-01-24</date><risdate>2022</risdate><abstract>The punctuation restoration task aims to correctly punctuate the output transcriptions of automatic speech recognition systems. Previous punctuation models, either using text only or demanding the corresponding audio, tend to be constrained by real scenes, where unpunctuated sentences are a mixture of those with and without audio. This paper proposes a unified multimodal punctuation restoration framework, named UniPunc, to punctuate the mixed sentences with a single model. UniPunc jointly represents audio and non-audio samples in a shared latent space, based on which the model learns a hybrid representation and punctuates both kinds of samples. We validate the effectiveness of the UniPunc on real-world datasets, which outperforms various strong baselines (e.g. BERT, MuSe) by at least 0.8 overall F1 scores, making a new state-of-the-art. Extensive experiments show that UniPunc's design is a pervasive solution: by grafting onto previous models, UniPunc enables them to punctuate on the mixed corpus. Our code is available at github.com/Yaoming95/UniPunc</abstract><doi>10.48550/arxiv.2202.00468</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2202.00468
ispartof
issn
language eng
recordid cdi_arxiv_primary_2202_00468
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Computation and Language
title Unified Multimodal Punctuation Restoration Framework for Mixed-Modality Corpus
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-18T00%3A00%3A58IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Unified%20Multimodal%20Punctuation%20Restoration%20Framework%20for%20Mixed-Modality%20Corpus&rft.au=Zhu,%20Yaoming&rft.date=2022-01-24&rft_id=info:doi/10.48550/arxiv.2202.00468&rft_dat=%3Carxiv_GOX%3E2202_00468%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true