Trust-Aware Benchmarking of GAN, VAE, and Diffusion Models for Synthetic Data in Image and Tabular Domains

Trust-Aware Benchmarking of GAN, VAE, and Diffusion Models for Synthetic Data in Image and Tabular Domains | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Trust-Aware Benchmarking of GAN, VAE, and Diffusion Models for Synthetic Data in Image and Tabular Domains Anil Kumar Shukla This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7649434/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Synthetic data generation using generative AI models offers a promising solution to challenges in data availability, privacy, and fairness. However, comparative insights across model families and data modalities remain limited, especially when trust-related dimensions such as fairness, privacy, and efficiency are considered alongside fidelity. This paper presents a trust-aware benchmarking study across both visual (CIFAR-10, MNIST) and tabular (Adult Income) domains, evaluating representative baselines—vanilla GAN/WGAN-GP, standard and β-VAE, DDPM and classifier-guided DDPM, and hybrids such as VAE-GAN and Latent Diffusion. Advanced models including StyleGAN2/3, BigGAN, CTGAN, Diffusion Transformers, and GigaGAN are acknowledged to situate the findings within the evolving landscape of foundation-scale generators. Performance is assessed using a multi-objective framework that integrates fidelity (FID, precision/recall), fairness (demographic parity), privacy leakage resistance, and computational efficiency. Results show that hybrid latent diffusion models achieve near-diffusion fidelity (FID 10.2 vs. 8.5 on CIFAR-10; 7.8 vs. 6.2 on MNIST) while reducing sampling time by over 70%. On tabular data, hybrids balance accuracy (84.7%) and fairness (0.93), whereas classical GANs and VAEs exhibit trade-offs between fidelity, efficiency, and fairness. To the best of our knowledge, this is the first study to benchmark GANs, VAEs, diffusion, and hybrid models across both image and tabular data using a unified, trust-aware evaluation framework. By providing reproducible, cross-domain comparisons, this work offers practical guidance for selecting and deploying generative models in trust-sensitive applications such as healthcare, finance, and autonomy. Synthetic Data Generative AI GANs Variational Autoencoders (VAEs) Diffusion Models Hybrid Architectures Data Privacy Fairness Computational Efficiency Trust-aware Benchmarking Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7649434","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":528383051,"identity":"93fc5a9e-0f44-4c2f-8955-9d5bce3d1fd1","order_by":0,"name":"Anil Kumar Shukla","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYDACCQZmhgcMCQz8IE5CAbFaEoBQsgGkxYAULQYHQDxitMjP7jE2SKhJszc+vzrxwwMDBnl+sQP4tRjcOWOckHAsJ3HbjbebJYAOM5w5O4GAFokc4wMJbBUJZjfObgBpSTC4TUCL/AyQln8V9sYzzm7+QZQWhhs5xgmJbTmMG/h7txFni8GNtGKDxL60xBk3eLdZJBhIEPaL/IzkzRIfviXb8/ef3XzzR4WNPL80IYfBgQRYpQSxykGA/wApqkfBKBgFo2AkAQCDV0VybS2MHgAAAABJRU5ErkJggg==","orcid":"","institution":"Principal – AI Labs Infosys Consulting","correspondingAuthor":true,"prefix":"","firstName":"Anil","middleName":"Kumar","lastName":"Shukla","suffix":""}],"badges":[],"createdAt":"2025-09-18 12:08:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7649434/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7649434/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93411344,"identity":"0fea08cb-c605-4171-8e38-7ea928415d9f","added_by":"auto","created_at":"2025-10-13 14:36:45","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":365872,"visible":true,"origin":"","legend":"","description":"","filename":"MainManuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/77f12f90a7c8d7218211f1b6.docx"},{"id":93409370,"identity":"f956c87e-eddc-44d6-b4e8-b9eb4f001b4e","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3834,"visible":true,"origin":"","legend":"","description":"","filename":"38b4a03b31d042389084a497965a0557.json","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/abd0eded3fb6ae0bde5a72bd.json"},{"id":93410798,"identity":"4e32967b-720b-49f0-89c0-653bc709cc52","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":173809,"visible":true,"origin":"","legend":"","description":"","filename":"38b4a03b31d042389084a497965a05571enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/e22aee62cbd48c21d77985b6.xml"},{"id":93409371,"identity":"dddbf04a-393f-400f-801a-929cf2740115","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6662,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/9c4dd4af895fd82ea3ebdb59.png"},{"id":93409376,"identity":"ef1ff042-7fd8-4584-ba8d-6df86f40ed4e","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36991,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/90017610b0caee84f7a508de.png"},{"id":93409372,"identity":"19b443fb-532b-4696-8426-1364737564f3","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6648,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/e278ac3baa6262d4d8ca7c22.png"},{"id":93411342,"identity":"e4501f56-b8e5-4278-893b-9a502d775e14","added_by":"auto","created_at":"2025-10-13 14:36:45","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37522,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/2972a9650884a57b6e313d20.png"},{"id":93412333,"identity":"f1247c4c-ddb3-4a6e-a87e-b53a90a33219","added_by":"auto","created_at":"2025-10-13 14:44:45","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3495,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/cf4689239017ec3e3e701ebf.png"},{"id":93409378,"identity":"d5892245-9246-4624-8800-c250a190d500","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4472,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/30608c71ecd0fbd266a264bb.png"},{"id":93410795,"identity":"1b13c4fd-0b61-4015-9264-7e1d790ec844","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":59985,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/1a547a011a5ceaf2de0d5f21.png"},{"id":93411347,"identity":"41e30749-8de0-4f43-bb0a-ad6e4e8cdc77","added_by":"auto","created_at":"2025-10-13 14:36:45","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":216785,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/506f5f31bac99f1ece1ba35c.jpeg"},{"id":93409389,"identity":"f36d98df-5a2f-43fb-8f7d-2b294b2cd7cd","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":262082,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/f8c83caab9a1522acdd4d1fe.jpeg"},{"id":93410800,"identity":"d1cc7f6e-f4c5-46b6-99ff-496a5ee432a2","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3595,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/e4dc37df34bdea30bf8573d0.png"},{"id":93409382,"identity":"5925f5c9-bb21-4144-a495-c1dc139ebfde","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11462,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/8277f3bd2db10c05331062b0.png"},{"id":93410805,"identity":"42e3dffb-ff55-453b-9e1c-bc0342f6154d","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3703,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/144a9f7a401be02f5624e6aa.png"},{"id":93410799,"identity":"635b6e3a-e1c4-450b-8563-b0a30cf4010a","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11266,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/315923ca22806f06aa2ce7dd.png"},{"id":93413100,"identity":"64a2edd6-3053-4cab-9251-9865054f7951","added_by":"auto","created_at":"2025-10-13 14:52:45","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2029,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/277c4fbc1c2ceee2d30ec7d2.png"},{"id":93412334,"identity":"0dcbfe12-b0e4-4629-842c-a2037c1705c7","added_by":"auto","created_at":"2025-10-13 14:44:45","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2473,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/01829846af70dff3000f2e23.png"},{"id":93409379,"identity":"bb3259e8-0598-4a52-aa11-f2361400ae8b","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17797,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/fb5f5f1d0b348b2df553ceec.png"},{"id":93409390,"identity":"b9ea0c42-a628-4210-92f4-5005657ef93f","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":57365,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/92e013d4d55f1b0f6d731980.png"},{"id":93409387,"identity":"d29db281-c2f5-4b94-8b65-06a332b23343","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":38067,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/8292b986b6566b6bc833a7a7.png"},{"id":93410806,"identity":"4aa4425b-2d0b-43b9-99f2-0f2643ddf566","added_by":"auto","created_at":"2025-10-13 14:28:45","extension":"xml","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":168081,"visible":true,"origin":"","legend":"","description":"","filename":"38b4a03b31d042389084a497965a05571structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/829334dbe82c4db9e9f602b4.xml"},{"id":93409391,"identity":"ed476188-8d53-46de-8562-f3c33aadf1cc","added_by":"auto","created_at":"2025-10-13 14:20:45","extension":"html","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":186581,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1/c292e5bd2cb6bd5f2d364826.html"},{"id":94987044,"identity":"bd9a1822-c4a3-4eba-a5f5-2cd4d19130c8","added_by":"auto","created_at":"2025-11-03 07:01:10","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":660043,"visible":true,"origin":"","legend":"","description":"","filename":"MainManuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7649434/v1_covered_e4d3d386-2961-4bf1-96d4-f1b532c85818.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Trust-Aware Benchmarking of GAN, VAE, and Diffusion Models for Synthetic Data in Image and Tabular Domains","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Synthetic Data, Generative AI, GANs, Variational Autoencoders (VAEs), Diffusion Models, Hybrid Architectures, Data Privacy, Fairness, Computational Efficiency, Trust-aware Benchmarking","lastPublishedDoi":"10.21203/rs.3.rs-7649434/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7649434/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSynthetic data generation using generative AI models offers a promising solution to challenges in data availability, privacy, and fairness. However, comparative insights across model families and data modalities remain limited, especially when trust-related dimensions such as fairness, privacy, and efficiency are considered alongside fidelity.\u003c/p\u003e\u003cp\u003eThis paper presents a trust-aware benchmarking study across both visual (CIFAR-10, MNIST) and tabular (Adult Income) domains, evaluating representative baselines\u0026mdash;vanilla GAN/WGAN-GP, standard and β-VAE, DDPM and classifier-guided DDPM, and hybrids such as VAE-GAN and Latent Diffusion. Advanced models including StyleGAN2/3, BigGAN, CTGAN, Diffusion Transformers, and GigaGAN are acknowledged to situate the findings within the evolving landscape of foundation-scale generators.\u003c/p\u003e\u003cp\u003ePerformance is assessed using a multi-objective framework that integrates fidelity (FID, precision/recall), fairness (demographic parity), privacy leakage resistance, and computational efficiency. Results show that hybrid latent diffusion models achieve near-diffusion fidelity (FID 10.2 vs. 8.5 on CIFAR-10; 7.8 vs. 6.2 on MNIST) while reducing sampling time by over 70%. On tabular data, hybrids balance accuracy (84.7%) and fairness (0.93), whereas classical GANs and VAEs exhibit trade-offs between fidelity, efficiency, and fairness.\u003c/p\u003e\u003cp\u003eTo the best of our knowledge, this is the first study to benchmark GANs, VAEs, diffusion, and hybrid models across both image and tabular data using a unified, trust-aware evaluation framework. By providing reproducible, cross-domain comparisons, this work offers practical guidance for selecting and deploying generative models in trust-sensitive applications such as healthcare, finance, and autonomy.\u003c/p\u003e","manuscriptTitle":"Trust-Aware Benchmarking of GAN, VAE, and Diffusion Models for Synthetic Data in Image and Tabular Domains","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-13 14:20:40","doi":"10.21203/rs.3.rs-7649434/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"40c9da71-2ddf-40dd-ad5a-9e33f1dbf2c1","owner":[],"postedDate":"October 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-31T18:23:31+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-13 14:20:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7649434","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7649434","identity":"rs-7649434","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}