Journal article · 2022 · Open access
Pre-trained Perceptual Features Improve Differentially Private Image Generation
Transactions on machine learning research · Published 25 May 2022 · DOI 10.48550/arxiv.2205.12900
In brief
“Pre-trained Perceptual Features Improve Differentially Private Image Generation” is a 2022 journal article by Fredrik Harder and 3 co-authors, published in Transactions on machine learning research. OpenAlex counts 24 citations as of 19 November 2026, which puts it in the top 8% of works published in 2022 by citations. It is open access under a CC BY licence: the full text is free to read at Transactions on machine learning research.
- Citations
- 24
- Most cited in
- 2024 (12)
- Field-weighted impact
- 1.35
- References
- 42
- Authors
- 4
Abstract
Training even moderately-sized generative models with differentially-private stochastic gradient descent (DP-SGD) is difficult: the required level of noise for reasonable levels of privacy is simply too large. We advocate instead building off a good, relevant representation on an informative public dataset, then learning to model the private data with that representation. In particular, we minimize the maximum mean discrepancy (MMD) between private target data and a generator's distribution, using a kernel based on perceptual features learned from a public dataset. With the MMD, we can simply privatize the data-dependent term once and for all, rather than introducing noise at each step of optimization as in DP-SGD. Our algorithm allows us to generate CIFAR10-level images with $ε\approx 2$ which capture distinctive features in the distribution, far surpassing the current state of the art, which mostly focuses on datasets such as MNIST and FashionMNIST at a large $ε\approx 10$. Our work introduces simple yet powerful foundations for reducing the gap between private and non-private deep generative models. Our code is available at \url{https://github.com/ParkLabML/DP-MEPF}.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Harder, F., Asadabadi, M. J., Sutherland, D. J., & Park, M. (2022). Pre-trained Perceptual Features Improve Differentially Private Image Generation. Transactions on machine learning research. https://doi.org/10.48550/arxiv.2205.12900
MLA 9
Harder, Fredrik, et al. “Pre-trained Perceptual Features Improve Differentially Private Image Generation.” Transactions on machine learning research, 2022. https://doi.org/10.48550/arxiv.2205.12900.
Chicago
Harder, Fredrik, Milad Jalali Asadabadi, Danica J. Sutherland, and Mijung Park. 2022. “Pre-trained Perceptual Features Improve Differentially Private Image Generation.” Transactions on machine learning research. https://doi.org/10.48550/arxiv.2205.12900.
Harvard
Harder, F. et al. (2022) ‘Pre-trained Perceptual Features Improve Differentially Private Image Generation’, Transactions on machine learning research. doi:10.48550/arxiv.2205.12900.
IEEE
F. Harder, M. J. Asadabadi, D. J. Sutherland, and M. Park, “Pre-trained Perceptual Features Improve Differentially Private Image Generation,” Transactions on machine learning research, 2022, doi: 10.48550/arxiv.2205.12900.
BibTeX
@article{harder2022trained,
title = {{Pre-trained Perceptual Features Improve Differentially Private Image Generation}},
author = {Harder, Fredrik and Asadabadi, Milad Jalali and Sutherland, Danica J. and Park, Mijung},
journal = {Transactions on machine learning research},
year = {2022},
doi = {10.48550/arxiv.2205.12900},
url = {https://doi.org/10.48550/arxiv.2205.12900}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Citations by year
| Year | Citations | Share of the best year |
|---|---|---|
| 2022 | 2 | |
| 2023 | 4 | |
| 2024 | 12 | |
| 2025 | 6 |
Its field-weighted citation impact is 1.35: it has been cited 1.35 times as often as the average work of the same field, type and year (1.00 is the world average).
Topic and keywords
Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- maximum mean discrepancy
- perceptual features
- pre-trained representations
- generative adversarial networks
- CIFAR-10
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Harder, F., Asadabadi, M. J., Sutherland, D. J., & Park, M. (2022). Pre-trained Perceptual Features Improve Differentially Private Image Generation. Transactions on machine learning research. https://doi.org/10.48550/arxiv.2205.12900
How many times has this paper been cited?
24 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at Transactions on machine learning research: sign in and open the full text from this page.
Who wrote this paper?
Fredrik Harder, Milad Jalali Asadabadi, Danica J. Sutherland, Mijung Park.
What is the DOI of this paper?
10.48550/arxiv.2205.12900, permanently at https://doi.org/10.48550/arxiv.2205.12900. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.