<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>2162</id>
    <completedYear>2023</completedYear>
    <publishedYear>2023</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst>943</pageFirst>
    <pageLast>948</pageLast>
    <pageNumber>943 - 948</pageNumber>
    <edition/>
    <issue/>
    <volume/>
    <type>conferenceobject</type>
    <publisherName>IEEE</publisherName>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>1</belongsToBibliography>
    <completedDate>--</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Generative Models for Improved Naturalness, Intelligibility, and Voicing of Whispered Speech</title>
    <abstract language="eng">This work adapts two recent architectures of generative models and evaluates their effectiveness for the conversion of whispered speech to normal speech. We incorporate the normal target speech into the training criterion of vector-quantized variational autoencoders (VQ-VAEs) and Mel-GANs, thereby conditioning the systems to recover voiced speech from whispered inputs. Objective and subjective quality measures indicate that both VQ-VAEs and MelGANs can be modified to perform the conversion task. We find that the proposed approaches significantly improve the Mel cepstral distortion (MCD) metric by at least 25% relative to a Disco-GAN baseline. Subjective listening tests suggest that the MelGAN-based system significantly improves naturalness, intelligibility, and voicing compared to the whispered input speech. A novel evaluation measure based on differences between latent speech representations also indicates that our MelGAN-based approach yields improvements relative to the baseline.</abstract>
    <parentTitle language="eng">2022 IEEE Spoken Language Technology Workshop (SLT)</parentTitle>
    <identifier type="doi">10.1109/SLT54892.2023.10022796</identifier>
    <identifier type="arxiv">2212.01775v2</identifier>
    <identifier type="isbn">979-8-3503-9690-4</identifier>
    <enrichment key="opus_doi_flag">true</enrichment>
    <enrichment key="opus_import_data">{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T00:59:10Z","timestamp":1707872350633},"reference-count":35,"publisher":"IEEE","license":[{"start":{"date-parts":[[2023,1,9]],"date-time":"2023-01-09T00:00:00Z","timestamp":1673222400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2023,1,9]],"date-time":"2023-01-09T00:00:00Z","timestamp":1673222400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2023,1,9]]},"DOI":"10.1109\/slt54892.2023.10022796","type":"proceedings-article","created":{"date-parts":[[2023,1,27]],"date-time":"2023-01-27T18:54:03Z","timestamp":1674845643000},"source":"Crossref","is-referenced-by-count":1,"title":["Generative Models for Improved Naturalness, Intelligibility, and Voicing of Whispered Speech"],"prefix":"10.1109","author":[{"given":"Dominik","family":"Wagner","sequence":"first","affiliation":[{"name":"Technische Hochschule N&amp;#x00FC;rnberg Georg Simon Ohm,Germany"}]},{"given":"Sebastian P.","family":"Bayerl","sequence":"additional","affiliation":[{"name":"Technische Hochschule N&amp;#x00FC;rnberg Georg Simon Ohm,Germany"}]},{"given":"Hector A. Cordourier","family":"Maruri","sequence":"additional","affiliation":[{"name":"Intel Labs"}]},{"given":"Tobias","family":"Bocklet","sequence":"additional","affiliation":[{"name":"Technische Hochschule N&amp;#x00FC;rnberg Georg Simon Ohm,Germany"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1017\/cbo9781139166621"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1017\/cbo9780511627743"},{"key":"ref3","volume-title":"Tracheoesophageal Speech. A Multidimensional Assessment of Voice Quality","author":"van As","year":"2001"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2005-611"},{"key":"ref5","article-title":"End-to-end whisper to natural speech conversion using modified transformer network","volume":"abs\/2004.09347","author":"Niranjan","year":"2021","journal-title":"ArXiv"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1487"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2940700"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.5555\/2969033.2969125"},{"key":"ref9","article-title":"Attention-guided generative adversarial network for whisper to normal speech conversion","volume":"vol. abs\/2111.01342","author":"Gao","year":"2021","journal-title":"ArXiv"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.21437\/IberSPEECH.2018-25"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2017-1428"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.23919\/EUSIPCO.2019.8902961"},{"key":"ref13","first-page":"1857","article-title":"Learning to discover cross-domain relations with generative adversarial networks","volume-title":"Proceedings of the 34th International Conference on Machine Learning","volume":"70","author":"Kim","year":"2017"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.244"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/JSTSP.2013.2283471"},{"key":"ref16","article-title":"Melgan: Generative adversarial networks for conditional waveform synthesis","volume":"32","author":"Kumar","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref17","article-title":"Neural discrete representation learning","volume":"30","author":"van den Oord","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TASLP.2019.2938863"},{"key":"ref19","volume-title":"Computational differences between whispered and non-whispered speech","author":"Lim","year":"2010"},{"key":"ref20","article-title":"Auto-encoding variational bayes","volume-title":"2nd International Conference on Learning Representations","author":"Kingma","year":"2014"},{"key":"ref21","article-title":"Estimating or propagating gradients through stochastic neurons for conditional computation","volume":"abs\/1308.3432","author":"Bengio","year":"2013","journal-title":"ArXiv"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683143"},{"key":"ref23","first-page":"8000","article-title":"The challenge of realistic music generation: Modelling raw audio at scale","volume-title":"Proceedings of the 32nd International Conference on Neural Information Processing Systems","author":"Dieleman","year":"2018"},{"key":"ref24","article-title":"Theory and experiments on vector quantized autoencoders","volume":"abs\/1805.1106","author":"Roy","year":"2018","journal-title":"ArXiv"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN48605.2020.9207145"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2010.5495701"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.21437\/Eurospeech.2001-111"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1196"},{"key":"ref29","first-page":"12449","article-title":"wav2vec 2.0: A framework for self-supervised learning of speech representations","volume":"33","author":"Baevski","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.21437\/interspeech.2021-703"},{"key":"ref31","first-page":"1780","article-title":"KSoF: The Kassel State of Fluency Dataset \u2013 A Therapy Centered Dataset of Stuttering","volume-title":"Proceedings of the Language Resources and Evaluation Conference","author":"Bayerl","year":"2022"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1016\/j.jvoice.2023.01.012"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.35111\/17gk-bn40"},{"key":"ref34","first-page":"2579","article-title":"Visualizing data using t-sne","volume":"9","author":"van der Maaten","year":"2008","journal-title":"Journal of Machine Learning Research"},{"key":"ref35","first-page":"1558","article-title":"Autoencoding beyond pixels using a learned similarity metric","volume-title":"Proceedings of the 33rd International Conference on International Conference on Machine Learning","volume":"48","author":"Larsen","year":"2016"}],"event":{"name":"2022 IEEE Spoken Language Technology Workshop (SLT)","location":"Doha, Qatar","start":{"date-parts":[[2023,1,9]]},"end":{"date-parts":[[2023,1,12]]}},"container-title":["2022 IEEE Spoken Language Technology Workshop (SLT)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/10022052\/10022330\/10022796.pdf?arnumber=10022796","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,2,13]],"date-time":"2024-02-13T08:07:37Z","timestamp":1707811657000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10022796\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,1,9]]},"references-count":35,"URL":"http:\/\/dx.doi.org\/10.1109\/slt54892.2023.10022796","relation":{},"subject":[],"published":{"date-parts":[[2023,1,9]]}}}</enrichment>
    <enrichment key="local_crossrefDocumentType">proceedings-article</enrichment>
    <enrichment key="local_crossrefLicence">https://doi.org/10.15223/policy-029</enrichment>
    <enrichment key="local_import_origin">crossref</enrichment>
    <enrichment key="local_doiImportPopulated">PersonAuthorFirstName_1,PersonAuthorLastName_1,PersonAuthorFirstName_2,PersonAuthorLastName_2,PersonAuthorFirstName_3,PersonAuthorLastName_3,PersonAuthorFirstName_4,PersonAuthorLastName_4,Enrichmentconference_title,Enrichmentconference_place,PublisherName,TitleMain_1,TitleParent_1,CompletedYear,Enrichmentlocal_crossrefLicence</enrichment>
    <enrichment key="conference_title">2022 IEEE Spoken Language Technology Workshop (SLT)</enrichment>
    <enrichment key="conference_place">Doha, Qatar</enrichment>
    <enrichment key="Reviewstatus">Begutachtet/Reviewed</enrichment>
    <enrichment key="opus.source">doi-import</enrichment>
    <enrichment key="ConferenceStatement">2022 IEEE Spoken Language Technology Workshop (SLT), 09-12 January 2023, Doha, Qatar</enrichment>
    <enrichment key="opus.doi.autoCreate">false</enrichment>
    <enrichment key="opus.urn.autoCreate">false</enrichment>
    <author>Dominik Wagner</author>
    <author>Sebastian P. Bayerl</author>
    <author>Hector A. Cordourier Maruri</author>
    <author>Tobias Bocklet</author>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>whispered speech</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>speech conversion</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>VAE</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>GAN</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>generative models</value>
    </subject>
    <collection role="institutes" number="">Fakultät Informatik</collection>
    <collection role="Forschungsschwerpunkt" number="5">Digitalisierung &amp; Künstliche Intelligenz</collection>
    <collection role="institutes" number="">Zentrum für Künstliche Intelligenz (KIZ)</collection>
  </doc>
</export-example>
