Quaternion Vector Quantized Variational Autoencoder
Vector quantized variational autoencoders, as variants of variational autoencoders, effectively capture discrete representations by quantizing continuous latent spaces and are widely used in generative tasks. However, these models still face limitations in handling complex image reconstruction, part...
Uloženo v:
| Vydáno v: | IEEE signal processing letters Ročník 32; s. 151 - 155 |
|---|---|
| Hlavní autoři: | , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
New York
IEEE
01.01.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Témata: | |
| ISSN: | 1070-9908, 1558-2361 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Vector quantized variational autoencoders, as variants of variational autoencoders, effectively capture discrete representations by quantizing continuous latent spaces and are widely used in generative tasks. However, these models still face limitations in handling complex image reconstruction, particularly in preserving high-quality details. Moreover, quaternion neural networks have shown unique advantages in handling multi-dimensional data, indicating that integrating quaternion approaches could potentially improve the performance of these autoencoders. To this end, we propose QVQ-VAE, a lightweight network in the quaternion domain that introduces a quaternion-based quantization layer and training strategy to improve reconstruction precision. By fully leveraging quaternion operations, QVQ-VAE reduces the number of model parameters, thereby lowering computational resource demands. Extensive evaluations on face and general object reconstruction tasks show that QVQ-VAE consistently outperforms existing methods while using significantly fewer parameters. |
|---|---|
| AbstractList | Vector quantized variational autoencoders, as variants of variational autoencoders, effectively capture discrete representations by quantizing continuous latent spaces and are widely used in generative tasks. However, these models still face limitations in handling complex image reconstruction, particularly in preserving high-quality details. Moreover, quaternion neural networks have shown unique advantages in handling multi-dimensional data, indicating that integrating quaternion approaches could potentially improve the performance of these autoencoders. To this end, we propose QVQ-VAE, a lightweight network in the quaternion domain that introduces a quaternion-based quantization layer and training strategy to improve reconstruction precision. By fully leveraging quaternion operations, QVQ-VAE reduces the number of model parameters, thereby lowering computational resource demands. Extensive evaluations on face and general object reconstruction tasks show that QVQ-VAE consistently outperforms existing methods while using significantly fewer parameters. |
| Author | Zhang, Yang Luo, Hui Liu, Xin Sun, Jian |
| Author_xml | – sequence: 1 givenname: Hui orcidid: 0009-0005-6189-5260 surname: Luo fullname: Luo, Hui email: mogu_hl@163.com organization: Faculty of Innovation Engineering, Macau University of Science and Technology, Macau, China – sequence: 2 givenname: Xin orcidid: 0000-0001-9646-4448 surname: Liu fullname: Liu, Xin email: xiliu@must.edu.mo organization: Faculty of Innovation Engineering, Macau University of Science and Technology, Macau, China – sequence: 3 givenname: Jian orcidid: 0009-0006-6485-0751 surname: Sun fullname: Sun, Jian email: jiansun97@163.com organization: Faculty of Innovation Engineering, Macau University of Science and Technology, Macau, China – sequence: 4 givenname: Yang orcidid: 0000-0002-0540-0893 surname: Zhang fullname: Zhang, Yang email: yang.zhang@umanitoba.ca organization: Department of Mathematics, University of Manitoba, Winnipeg, MB, Canada |
| BookMark | eNpNkDtPwzAQgC1UJNrCzsAQiTnhHNuxPVYVLykSIKCr5SQXKVWJi-0M8Otx1Q5Md7r77qFvQWajG5GQawoFpaDv6vfXooSSF0wAZ5KfkTkVQuUlq-gs5SAh1xrUBVmEsAUARZWYE_Y22Yh-HNyYbbCNzmepMsbhF7tsY_1gY2rZXbaaosOxdR36S3Le213Aq1Ncks-H-4_1U16_PD6vV3XellzEvNNMSkUBhaZS9qqptG2E7vrW6nS_qSQwzhveqxJpqRjlrGlsJ3rdWYm0Yktye9y79-57whDN1k0-PRNMgiuhlKQqUXCkWu9C8NibvR--rP8xFMxBjUlqzEGNOalJIzfHkQER_-GyYlxw9gfRt2BJ |
| CODEN | ISPLEM |
| Cites_doi | 10.1109/CVPR.2009.5206848 10.1109/TPAMI.2023.3261988 10.1109/MSP.2023.3278071 10.1145/3626235 10.1109/ICASSP.2019.8682495 10.1007/978-3-030-56694-4 10.1007/978-3-030-68787-8_12 10.1016/j.sigpro.2010.06.024 10.1016/j.asoc.2024.112154 10.21437/Interspeech.2018-1898 10.3390/info12030117 10.1109/IJCNN.2018.8489651 10.1109/TPAMI.2024.3356232 10.1109/TCSS.2023.3305616 10.1109/CVPR.2019.00453 10.1109/TCSVT.2023.3283289 10.1109/CVPR.2016.308 10.1007/s00006-024-01350-x 10.1109/CVPRW63382.2024.00750 10.1109/LSP.2021.3135803 10.1137/0330046 10.1007/s11831-023-09925-w 10.1109/ICPR56361.2022.9956102 10.1016/S0893-6080(96)00048-2 10.1007/s11263-010-0388-9 10.1109/ICCV.2015.425 10.1109/ICASSP39728.2021.9413859 10.1109/TNN.2011.2157358 10.1145/3592450 10.1007/978-3-030-01237-3_39 10.1007/978-3-030-91390-8_4 |
| ContentType | Journal Article |
| Copyright | Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025 |
| Copyright_xml | – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025 |
| DBID | 97E RIA RIE AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D |
| DOI | 10.1109/LSP.2024.3504374 |
| DatabaseName | IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional |
| DatabaseTitle | CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional |
| DatabaseTitleList | Technology Research Database |
| Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering |
| EISSN | 1558-2361 |
| EndPage | 155 |
| ExternalDocumentID | 10_1109_LSP_2024_3504374 10763454 |
| Genre | orig-research |
| GrantInformation_xml | – fundername: Natural Sciences and Engineering Research Council of Canada; NSERC grantid: RGPIN 2020-06746 funderid: 10.13039/501100000038 – fundername: National Natural Science Foundation of China; NSFC grantid: 12371023; 12271338 funderid: 10.13039/501100001809 – fundername: Wuyi University, Hong Kong and Macao grantid: 2019WGALH20 |
| GroupedDBID | -~X .DC 0R~ 29I 3EH 4.4 5GY 5VS 6IK 85S 97E AAJGR AARMG AASAJ AAWTH AAYJJ ABAZT ABFSI ABQJQ ABVLG ACGFO ACGFS ACIWK AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 E.L EBS EJD F5P HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 O9- OCL P2P RIA RIE RNS TAE TN5 VH1 AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D |
| ID | FETCH-LOGICAL-c245t-d9377810e59177f8b69ab59dfca9000b670344b4f82e1283143bbad5f9da7e163 |
| IEDL.DBID | RIE |
| ISICitedReferencesCount | 0 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001380677100038&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 1070-9908 |
| IngestDate | Mon Jun 30 10:13:35 EDT 2025 Sat Nov 29 03:38:57 EST 2025 Wed Aug 27 01:57:01 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Language | English |
| License | https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037 |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c245t-d9377810e59177f8b69ab59dfca9000b670344b4f82e1283143bbad5f9da7e163 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ORCID | 0009-0005-6189-5260 0009-0006-6485-0751 0000-0001-9646-4448 0000-0002-0540-0893 |
| PQID | 3146588718 |
| PQPubID | 75747 |
| PageCount | 5 |
| ParticipantIDs | proquest_journals_3146588718 crossref_primary_10_1109_LSP_2024_3504374 ieee_primary_10763454 |
| PublicationCentury | 2000 |
| PublicationDate | 2025-01-01 |
| PublicationDateYYYYMMDD | 2025-01-01 |
| PublicationDate_xml | – month: 01 year: 2025 text: 2025-01-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationPlace | New York |
| PublicationPlace_xml | – name: New York |
| PublicationTitle | IEEE signal processing letters |
| PublicationTitleAbbrev | LSP |
| PublicationYear | 2025 |
| Publisher | IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Publisher_xml | – name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| References | ref35 ref12 ref15 ref14 ref31 ref30 ref11 ref33 ref10 ref32 ref2 ref17 ref16 ref19 ref18 Oord (ref4) 2017; 30 Krizhevsky (ref34) 2009 ref24 ref23 ref26 ref25 ref20 ref22 ref21 ref28 ref27 ref29 ref8 ref7 ref9 Parcollet (ref13) 2018 ref3 ref6 ref5 Kingma (ref1) 2013 |
| References_xml | – ident: ref35 doi: 10.1109/CVPR.2009.5206848 – ident: ref3 doi: 10.1109/TPAMI.2023.3261988 – ident: ref9 doi: 10.1109/MSP.2023.3278071 – ident: ref2 doi: 10.1145/3626235 – ident: ref26 doi: 10.1109/ICASSP.2019.8682495 – ident: ref24 doi: 10.1007/978-3-030-56694-4 – year: 2018 ident: ref13 article-title: Quaternion recurrent neural networks – ident: ref16 doi: 10.1007/978-3-030-68787-8_12 – ident: ref25 doi: 10.1016/j.sigpro.2010.06.024 – ident: ref19 doi: 10.1016/j.asoc.2024.112154 – ident: ref11 doi: 10.21437/Interspeech.2018-1898 – ident: ref14 doi: 10.3390/info12030117 – ident: ref28 doi: 10.1109/IJCNN.2018.8489651 – ident: ref6 doi: 10.1109/TPAMI.2024.3356232 – ident: ref20 doi: 10.1109/TCSS.2023.3305616 – year: 2009 ident: ref34 article-title: Learning multiple layers of features from tiny images – ident: ref33 doi: 10.1109/CVPR.2019.00453 – ident: ref17 doi: 10.1109/TCSVT.2023.3283289 – ident: ref31 doi: 10.1109/CVPR.2016.308 – ident: ref21 doi: 10.1007/s00006-024-01350-x – ident: ref8 doi: 10.1109/CVPRW63382.2024.00750 – ident: ref18 doi: 10.1109/LSP.2021.3135803 – ident: ref30 doi: 10.1137/0330046 – ident: ref27 doi: 10.1007/s11831-023-09925-w – ident: ref7 doi: 10.1109/ICPR56361.2022.9956102 – ident: ref10 doi: 10.1016/S0893-6080(96)00048-2 – ident: ref22 doi: 10.1007/s11263-010-0388-9 – ident: ref32 doi: 10.1109/ICCV.2015.425 – ident: ref23 doi: 10.1109/ICASSP39728.2021.9413859 – ident: ref29 doi: 10.1109/TNN.2011.2157358 – year: 2013 ident: ref1 article-title: Auto-encoding variational Bayes – volume: 30 start-page: 6306 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2017 ident: ref4 article-title: Neural discrete representation learning – ident: ref5 doi: 10.1145/3592450 – ident: ref12 doi: 10.1007/978-3-030-01237-3_39 – ident: ref15 doi: 10.1007/978-3-030-91390-8_4 |
| SSID | ssj0008185 |
| Score | 2.438671 |
| Snippet | Vector quantized variational autoencoders, as variants of variational autoencoders, effectively capture discrete representations by quantizing continuous... |
| SourceID | proquest crossref ieee |
| SourceType | Aggregation Database Index Database Publisher |
| StartPage | 151 |
| SubjectTerms | Convolution Decoding Face recognition Image quality Image reconstruction Indexes Multidimensional data Neural networks Parameters Quantization (signal) quaternion generative models quaternion neural networks Quaternions Task complexity Training vector quantized variational autoencoder Vectors |
| Title | Quaternion Vector Quantized Variational Autoencoder |
| URI | https://ieeexplore.ieee.org/document/10763454 https://www.proquest.com/docview/3146588718 |
| Volume | 32 |
| WOSCitedRecordID | wos001380677100038&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-2361 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0008185 issn: 1070-9908 databaseCode: RIE dateStart: 19940101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07T8MwED5BxQADzyIKBWVgYUhxndhOxgpRMVRVEVB1i-z4InVpUZsw8Os5OykqQgxsUZRI1ne-p33fAdzmTBdMYT-0QugwRomhMUgCiRyZljRCe57u6UiNx8lslk6aZnXfC4OI_vIZ9tyjP8u3y7xypTLScNKGWMS7sKuUqpu1vs2u8zz1BUMWkolNNmeSLL0fvUwoE-RxL3J8XSr-4YP8UJVflti7l-HRPxd2DIdNHBkMasGfwA4uTuFgi13wDKLnSrtyHyEfTH1xPqA3i3L-iTaYUo7c1AGDQVUuHZ-lxVUb3oaPrw9PYTMjIcx5LMrQUnihkj5DQXmXKhIjU21Eaotcu3GgRirH6WfiIuFIriii8MgYbUWRWk0iktE5tBbLBV5AgKnrm9XKWo6xptSV9XPOUfJcskIK7MDdBrXsvabCyHwKwdKMEM4cwlmDcAfaDqWt72qAOtDd4Jw1yrLOaFEUB1Hmllz-8dsV7HM3d9eXPrrQKlcVXsNe_lHO16sbvw--APYFr2E |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LS8NAEB60CurBZ8Vq1Ry8eEhNN7t5HItYKtZSsZbewm52Ar200iYe_PXOblKpiAdvISSwfLPz3J1vAG5ST2ZeiG1XCyFdjgG6SiEJxDdkWoES0vJ0j_vhYBBNJvGwala3vTCIaC-fYcs82rN8PU8LUyojDSdt4IJvwpbgnLXLdq1vw2t8T3nF0HPJyEarU0kvvuu_DikXZLzlG8aukP_wQnasyi9bbB1M9-CfSzuE_SqSdDql6I9gA2fHsLfGL3gC_kshTcGPsHfGtjzv0JtZPv1E7YwpS64qgU6nyOeG0VLjog5v3YfRfc-tpiS4KeMidzUFGGHU9lBQ5hVmkQpiqUSss1SagaAqCA2rn-JZxJCckU8BklJSiyzWkoQU-KdQm81neAYOxqZzVoZaM-SSklevnTKGAUsDLwsENuB2hVryXpJhJDaJ8OKEEE4MwkmFcAPqBqW170qAGtBc4ZxU6rJMaFEUCVHuFp3_8ds17PRGz_2k_zh4uoBdZqbw2kJIE2r5osBL2E4_8ulycWX3xBdHJ7Ko |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Quaternion+Vector+Quantized+Variational+Autoencoder&rft.jtitle=IEEE+signal+processing+letters&rft.au=Luo%2C+Hui&rft.au=Liu%2C+Xin&rft.au=Sun%2C+Jian&rft.au=Zhang%2C+Yang&rft.date=2025-01-01&rft.issn=1070-9908&rft.eissn=1558-2361&rft.volume=32&rft.spage=151&rft.epage=155&rft_id=info:doi/10.1109%2FLSP.2024.3504374&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_LSP_2024_3504374 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1070-9908&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1070-9908&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1070-9908&client=summon |