{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,3,11]],"date-time":"2025-03-11T04:27:13Z","timestamp":1741667233238,"version":"3.38.0"},"reference-count":71,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"6","license":[{"start":{"date-parts":[[2025,3,15]],"date-time":"2025-03-15T00:00:00Z","timestamp":1741996800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,3,15]],"date-time":"2025-03-15T00:00:00Z","timestamp":1741996800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,3,15]],"date-time":"2025-03-15T00:00:00Z","timestamp":1741996800000},"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":["IEEE Internet Things J."],"published-print":{"date-parts":[[2025,3,15]]},"DOI":"10.1109\/jiot.2024.3496920","type":"journal-article","created":{"date-parts":[[2024,11,13]],"date-time":"2024-11-13T18:58:08Z","timestamp":1731524288000},"page":"7549-7563","source":"Crossref","is-referenced-by-count":0,"title":["DPAF: Image Synthesis via Differentially Private Aggregation in Forward Phase"],"prefix":"10.1109","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-2668-0556","authenticated-orcid":false,"given":"Chih-Hsun","family":"Lin","sequence":"first","affiliation":[{"name":"Department of Computer Science, National Yang Ming Chiao Tung University, Hsinchu, Taiwan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-7772-8735","authenticated-orcid":false,"given":"Chia-Yi","family":"Hsu","sequence":"additional","affiliation":[{"name":"Department of Computer Science, National Yang Ming Chiao Tung University, Hsinchu, Taiwan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1677-2131","authenticated-orcid":false,"given":"Chia-Mu","family":"Yu","sequence":"additional","affiliation":[{"name":"Institute of Electrical and Computer Engineering, National Yang Ming Chiao Tung University, Hsinchu, Taiwan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6424-8633","authenticated-orcid":false,"given":"Yang","family":"Cao","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Institute of Science Tokyo, Tokyo, Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5503-9541","authenticated-orcid":false,"given":"Chun-Ying","family":"Huang","sequence":"additional","affiliation":[{"name":"Department of Computer Science, National Yang Ming Chiao Tung University, Hsinchu, Taiwan"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/3372297.3417238"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.2478\/popets-2019-0008"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.2478\/popets-2019-0067"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1561\/9781601988195"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/2976749.2978318"},{"key":"ref6","first-page":"1","article-title":"Differential privacy has disparate impact on model accuracy","volume-title":"Proc. Conf. Neural Inf. Process. Syst. (NeurIPS)","author":"Bagdasaryan"},{"key":"ref7","first-page":"1","article-title":"GS-WGAN: A gradient-sanitized approach for learning differentially private generators","volume-title":"Proc. Conf. Neural Inf. Process. Syst. (NeurIPS)","author":"Chen"},{"key":"ref8","first-page":"214","article-title":"Wasserstein generative adversarial networks","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Arjovsky"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3460120.3484579"},{"key":"ref10","first-page":"929","article-title":"PrivSyn: Differentially private data synthesis","volume-title":"Proc. USENIX Security Symp.","author":"Zhang"},{"key":"ref11","first-page":"4435","article-title":"Graphical-model based estimation and inference for differential privacy","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Mckenna"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3134428"},{"key":"ref13","article-title":"Unlocking high-accuracy differentially private image classification through scale","author":"De","year":"2022","journal-title":"arXiv: 2204.13650"},{"key":"ref14","first-page":"1","article-title":"Differentially private learning needs better features (or much more data)","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Tram\u00e8r"},{"key":"ref15","first-page":"1","article-title":"Large scale private learning via low-rank reparametrization","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Yu"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.29012\/jpc.880"},{"key":"ref17","first-page":"1","article-title":"Large language models can be strong differentially private learners","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Li"},{"key":"ref18","first-page":"1","article-title":"Semi-supervised knowledge transfer for deep learning from private training data","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Papernot"},{"key":"ref19","first-page":"1","article-title":"Scalable private learning with PATE","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Papernot"},{"key":"ref20","first-page":"1","article-title":"PATE-GAN: Generating synthetic data with differential privacy guarantees","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Jordon"},{"key":"ref21","article-title":"Differentially private releasing via deep generative model","author":"Zhang","year":"2018","journal-title":"arXiv: 1801.01594"},{"key":"ref22","first-page":"1","article-title":"A general approach to adding differential privacy to iterative training procedures","volume-title":"Proc. NeurIPS Workshop Privacy Preserving Mach. Learn. (PPML)","author":"McMahan"},{"key":"ref23","first-page":"1","article-title":"Differentially private learning with adaptive clipping","volume-title":"Proc. Conf. Neural Inf. Process. Syst. (NeurIPS)","author":"Thakkar"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/TIFS.2019.2897874"},{"key":"ref25","article-title":"Differentially private generative adversarial network","author":"Xie","year":"2018","journal-title":"arXiv:1802.06739"},{"key":"ref26","first-page":"1","article-title":"Do not let privacy overbill utility: Gradient embedding perturbation for private learning","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Yu"},{"key":"ref27","first-page":"1","article-title":"Improving deep learning with differential privacy using gradient encoding and denoising","volume-title":"Proc. Theory Pract. Differ. Privacy","author":"Nasr"},{"key":"ref28","first-page":"1","article-title":"G-PATE: Scalable differentially private data generator via private aggregation of teacher discriminators","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Long"},{"key":"ref29","first-page":"1","article-title":"DP-MERF: Differentially private mean embeddings with random features for practical privacy-preserving data generation","volume-title":"Proc. Int. Conf. Artif. Intell. Stat. (AISTATS)","author":"Harder"},{"journal-title":"Trans. Mach. Learn. Res.","article-title":"Pretrained perceptual features improve differentially private image generation","author":"Harder","key":"ref30"},{"key":"ref31","first-page":"1","article-title":"Differentially private neural tangent kernels for privacy-preserving data generation","volume-title":"Proc. AAAI Workshop Privacy-Preserving Artif. Intell. (PPAI)","author":"Yang"},{"key":"ref32","first-page":"22300","article-title":"Hermite polynomial features for private data generation","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Vinaroz"},{"key":"ref33","first-page":"1","article-title":"Don\u2019t generate me: Training differentially private generative models with Sinkhorn divergence","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Cao"},{"key":"ref34","first-page":"1","article-title":"PEARL: Data synthesis via private embeddings and adversarial reconstruction learning","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Liew"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/ICDE51399.2021.00022"},{"key":"ref36","article-title":"DP2-VAE: Differentially private pretrained variational autoencoders","author":"Jiang","year":"2022","journal-title":"arXiv:2208.03409"},{"key":"ref37","article-title":"DPD-fVAE: Synthetic data generation using federated variational autoencoders with differentially-private decoder","author":"Pfitzner","year":"2022","journal-title":"arXiv:2211.11591"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00820"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.14722\/aiscc.2024.23005"},{"key":"ref40","first-page":"1","article-title":"Dataset condensation with gradient matching","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Zhao"},{"key":"ref41","first-page":"1","article-title":"Differentially private diffusion models","volume-title":"Proc. Trans. Mach. Learn. Res.","author":"Dockhorn"},{"key":"ref42","first-page":"1","article-title":"Diffusion models beat GANs on image synthesis","author":"Dhariwal","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst. (NeurIPS)"},{"key":"ref43","first-page":"1","article-title":"Denoising diffusion probabilistic models","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Ho"},{"key":"ref44","first-page":"1","article-title":"Differentially private diffusion models generate useful synthetic images","volume-title":"Proc. Int. Workshop Trustworthy Federated Learn.","author":"Ghalebikesabi"},{"key":"ref45","first-page":"1063","article-title":"DP-promise: Differentially private diffusion probabilistic models for image synthesis","volume-title":"Proc. USENIX Security Symp.","author":"Wang"},{"key":"ref46","first-page":"4837","article-title":"{PrivImage}: Differentially private synthetic image generation using diffusion models with {semantic-aware} pretraining","volume-title":"Proc. 33rd USENIX Security Symp. (USENIX Security)","author":"Li"},{"key":"ref47","first-page":"1","article-title":"Differentially private synthetic data via foundation model APIs 1: Images","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Lin"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1145\/1989323.1989345"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.56553\/popets-2022-0122"},{"key":"ref50","article-title":"Conditional generative adversarial nets","author":"Mirza","year":"2014","journal-title":"arXiv:1411.1784"},{"key":"ref51","first-page":"2642","article-title":"Conditional image synthesis with auxiliary classifier GANs","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Odena"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1511.06434"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2017.41"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1145\/3548606.3560675"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1145\/2810103.2813677"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/SP40001.2021.00099"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.437"},{"key":"ref58","first-page":"1","volume-title":"Instance normalization: The missing ingredient for fast stylization","author":"Ulyanov","year":"2016"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr42600.2020.00321"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.167"},{"key":"ref61","first-page":"1","article-title":"Invertible conditional GANs for image editing","volume-title":"Proc. NeurIPS Workshop Adversarial Training","author":"Perarnau"},{"key":"ref62","first-page":"1","article-title":"Deep gradient compression: Reducing the communication bandwidth for distributed training","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Lin"},{"key":"ref63","first-page":"1","article-title":"Demystifying MMD GANs","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Bi\u0144kowski"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-020-01424-w"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref66","first-page":"1","article-title":"Progressive growing of GANs for improved quality, stability, and variation","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Karras"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/FOCS.2014.56"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i10.17123"},{"key":"ref69","first-page":"1","article-title":"Private GANs, revisited","volume-title":"Proc. Trans. Mach. Learn. Res.","author":"Bie"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/CSF.2017.11"},{"key":"ref71","first-page":"1226","article-title":"Subsampled R\u00e9nyi differential privacy and analytical moments accountant","volume-title":"Proc. Int. Conf. Artif. Intell. Stat. (AISTATS)","author":"Wang"}],"container-title":["IEEE Internet of Things Journal"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/6488907\/10918322\/10752538.pdf?arnumber=10752538","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,10]],"date-time":"2025-03-10T17:45:07Z","timestamp":1741628707000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10752538\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,15]]},"references-count":71,"journal-issue":{"issue":"6"},"URL":"https:\/\/doi.org\/10.1109\/jiot.2024.3496920","relation":{},"ISSN":["2327-4662","2372-2541"],"issn-type":[{"type":"electronic","value":"2327-4662"},{"type":"electronic","value":"2372-2541"}],"subject":[],"published":{"date-parts":[[2025,3,15]]}}}