ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis

Representing speech as discretized units has numerous benefits in supporting downstream spoken language processing tasks. However, the approach has been less explored in speech synthesis of tonal languages like Mandarin Chinese. Our preliminary experiments on Chinese speech synthesis reveal the issu...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tao, Dehua, Tan, Daxin, Yeung, Yu Ting, Chen, Xiao, Lee, Tan
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 Tao, Dehua
Tan, Daxin
Yeung, Yu Ting
Chen, Xiao
Lee, Tan
description Representing speech as discretized units has numerous benefits in supporting downstream spoken language processing tasks. However, the approach has been less explored in speech synthesis of tonal languages like Mandarin Chinese. Our preliminary experiments on Chinese speech synthesis reveal the issue of "tone shift", where a synthesized speech utterance contains correct base syllables but incorrect tones. To address the issue, we propose the ToneUnit framework, which leverages annotated data with tone labels as CTC supervision to learn tone-aware discrete speech units for Mandarin Chinese speech. Our findings indicate that the discrete units acquired through the TonUnit resolve the "tone shift" issue in synthesized Chinese speech and yield favorable results in English synthesis. Moreover, the experimental results suggest that finite scalar quantization enhances the effectiveness of ToneUnit. Notably, ToneUnit can work effectively even with minimal annotated data.
doi_str_mv 10.48550/arxiv.2406.08989
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2406_08989</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2406_08989</sourcerecordid><originalsourceid>FETCH-LOGICAL-a679-2f5f9930b9b28dd18cd11c17b5e845c13bc06c1aa790b323643699261cadef423</originalsourceid><addsrcrecordid>eNo1j8tOwzAURL1hgQofwAr_QILfsbuLylOKYNF0HV07dmupOJETEOXrKS2sRhrNGekgdENJKbSU5A7yV_wsmSCqJNpoc4le2yH5TYrzEtd4PXrvdvg-Ti77OX7DHIeE63HMAxz7MGR8nMMeN5C2H7D1_8T6kOadn-J0hS4C7Cd__ZcL1D4-tKvnonl7elnVTQGqMgULMhjDiTWW6b6n2vWUOlpZ6bWQjnLriHIUoDLEcsaV4MoYpqiD3gfB-ALdnm9PQt2Y4zvkQ_cr1p3E-A9Xm0hs</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis</title><source>arXiv.org</source><creator>Tao, Dehua ; Tan, Daxin ; Yeung, Yu Ting ; Chen, Xiao ; Lee, Tan</creator><creatorcontrib>Tao, Dehua ; Tan, Daxin ; Yeung, Yu Ting ; Chen, Xiao ; Lee, Tan</creatorcontrib><description>Representing speech as discretized units has numerous benefits in supporting downstream spoken language processing tasks. However, the approach has been less explored in speech synthesis of tonal languages like Mandarin Chinese. Our preliminary experiments on Chinese speech synthesis reveal the issue of "tone shift", where a synthesized speech utterance contains correct base syllables but incorrect tones. To address the issue, we propose the ToneUnit framework, which leverages annotated data with tone labels as CTC supervision to learn tone-aware discrete speech units for Mandarin Chinese speech. Our findings indicate that the discrete units acquired through the TonUnit resolve the "tone shift" issue in synthesized Chinese speech and yield favorable results in English synthesis. Moreover, the experimental results suggest that finite scalar quantization enhances the effectiveness of ToneUnit. Notably, ToneUnit can work effectively even with minimal annotated data.</description><identifier>DOI: 10.48550/arxiv.2406.08989</identifier><language>eng</language><subject>Computer Science - Sound</subject><creationdate>2024-06</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>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2406.08989$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2406.08989$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Tao, Dehua</creatorcontrib><creatorcontrib>Tan, Daxin</creatorcontrib><creatorcontrib>Yeung, Yu Ting</creatorcontrib><creatorcontrib>Chen, Xiao</creatorcontrib><creatorcontrib>Lee, Tan</creatorcontrib><title>ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis</title><description>Representing speech as discretized units has numerous benefits in supporting downstream spoken language processing tasks. However, the approach has been less explored in speech synthesis of tonal languages like Mandarin Chinese. Our preliminary experiments on Chinese speech synthesis reveal the issue of "tone shift", where a synthesized speech utterance contains correct base syllables but incorrect tones. To address the issue, we propose the ToneUnit framework, which leverages annotated data with tone labels as CTC supervision to learn tone-aware discrete speech units for Mandarin Chinese speech. Our findings indicate that the discrete units acquired through the TonUnit resolve the "tone shift" issue in synthesized Chinese speech and yield favorable results in English synthesis. Moreover, the experimental results suggest that finite scalar quantization enhances the effectiveness of ToneUnit. Notably, ToneUnit can work effectively even with minimal annotated data.</description><subject>Computer Science - Sound</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNo1j8tOwzAURL1hgQofwAr_QILfsbuLylOKYNF0HV07dmupOJETEOXrKS2sRhrNGekgdENJKbSU5A7yV_wsmSCqJNpoc4le2yH5TYrzEtd4PXrvdvg-Ti77OX7DHIeE63HMAxz7MGR8nMMeN5C2H7D1_8T6kOadn-J0hS4C7Cd__ZcL1D4-tKvnonl7elnVTQGqMgULMhjDiTWW6b6n2vWUOlpZ6bWQjnLriHIUoDLEcsaV4MoYpqiD3gfB-ALdnm9PQt2Y4zvkQ_cr1p3E-A9Xm0hs</recordid><startdate>20240613</startdate><enddate>20240613</enddate><creator>Tao, Dehua</creator><creator>Tan, Daxin</creator><creator>Yeung, Yu Ting</creator><creator>Chen, Xiao</creator><creator>Lee, Tan</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240613</creationdate><title>ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis</title><author>Tao, Dehua ; Tan, Daxin ; Yeung, Yu Ting ; Chen, Xiao ; Lee, Tan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a679-2f5f9930b9b28dd18cd11c17b5e845c13bc06c1aa790b323643699261cadef423</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Sound</topic><toplevel>online_resources</toplevel><creatorcontrib>Tao, Dehua</creatorcontrib><creatorcontrib>Tan, Daxin</creatorcontrib><creatorcontrib>Yeung, Yu Ting</creatorcontrib><creatorcontrib>Chen, Xiao</creatorcontrib><creatorcontrib>Lee, Tan</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Tao, Dehua</au><au>Tan, Daxin</au><au>Yeung, Yu Ting</au><au>Chen, Xiao</au><au>Lee, Tan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis</atitle><date>2024-06-13</date><risdate>2024</risdate><abstract>Representing speech as discretized units has numerous benefits in supporting downstream spoken language processing tasks. However, the approach has been less explored in speech synthesis of tonal languages like Mandarin Chinese. Our preliminary experiments on Chinese speech synthesis reveal the issue of "tone shift", where a synthesized speech utterance contains correct base syllables but incorrect tones. To address the issue, we propose the ToneUnit framework, which leverages annotated data with tone labels as CTC supervision to learn tone-aware discrete speech units for Mandarin Chinese speech. Our findings indicate that the discrete units acquired through the TonUnit resolve the "tone shift" issue in synthesized Chinese speech and yield favorable results in English synthesis. Moreover, the experimental results suggest that finite scalar quantization enhances the effectiveness of ToneUnit. Notably, ToneUnit can work effectively even with minimal annotated data.</abstract><doi>10.48550/arxiv.2406.08989</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2406.08989
ispartof
issn
language eng
recordid cdi_arxiv_primary_2406_08989
source arXiv.org
subjects Computer Science - Sound
title ToneUnit: A Speech Discretization Approach for Tonal Language Speech Synthesis
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-02T23%3A49%3A38IST&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=ToneUnit:%20A%20Speech%20Discretization%20Approach%20for%20Tonal%20Language%20Speech%20Synthesis&rft.au=Tao,%20Dehua&rft.date=2024-06-13&rft_id=info:doi/10.48550/arxiv.2406.08989&rft_dat=%3Carxiv_GOX%3E2406_08989%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