# DPImageBench: A Unified Benchmark for Differentially Private Image Synthesis

arXiv: https://arxiv.org/abs/2503.14681 · HTML: https://mlsanity.com/paper/2503.14681/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 62 of its 94 references that are in the map.
- First posted on arXiv 18 Mar 2025.

## Builds on (most cited first, 30 of 62)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Deep Learning Face Attributes in the Wild](https://mlsanity.com/paper/1411.7766/) (2014)
- [Wide Residual Networks](https://mlsanity.com/paper/1605.07146/) (2016)
- [Elucidating the Design Space of Diffusion-Based Generative Models](https://mlsanity.com/paper/2206.00364/) (2022)
- [Improved Techniques for Training GANs](https://mlsanity.com/paper/1606.03498/) (2016)
- [Deep Learning with Differential Privacy](https://mlsanity.com/paper/1607.00133/) (2016)
- [Large Scale GAN Training for High Fidelity Natural Image Synthesis](https://mlsanity.com/paper/1809.11096/) (2018)
- [Aggregated Residual Transformations for Deep Neural Networks](https://mlsanity.com/paper/1611.05431/) (2016)
- [EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification](https://mlsanity.com/paper/1709.00029/) (2017)
- [Show, Attend and Tell: Neural Image Caption Generation with Visual Attention](https://mlsanity.com/paper/1502.03044/) (2015)
- [ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2304.05977/) (2023)
- [Improved Precision and Recall Metric for Assessing Generative Models](https://mlsanity.com/paper/1904.06991/) (2019)
- [Show and Tell: A Neural Image Caption Generator](https://mlsanity.com/paper/1411.4555/) (2014)
- [Renyi Differential Privacy](https://mlsanity.com/paper/1702.07476/) (2017)
- [On Exact Computation with an Infinitely Wide Neural Net](https://mlsanity.com/paper/1904.11955/) (2019)
- [NIPS 2016 Tutorial: Generative Adversarial Networks](https://mlsanity.com/paper/1701.00160/) (2016)
- [Differentially Private Generative Adversarial Network](https://mlsanity.com/paper/1802.06739/) (2018)
- [Unlocking High-Accuracy Differentially Private Image Classification through Scale](https://mlsanity.com/paper/2204.13650/) (2022)
- [Auditing Differentially Private Machine Learning: How Private is Private SGD?](https://mlsanity.com/paper/2006.07709/) (2020)
- [R'enyi Differential Privacy of the Sampled Gaussian Mechanism](https://mlsanity.com/paper/1908.10530/) (2019)
- [VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learning](https://mlsanity.com/paper/1705.07761/) (2017)
- [Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning](https://mlsanity.com/paper/2101.04535/) (2021)

## Built on by (most cited first, 11 of 11)

- [Synthesize Privacy-Preserving High-Resolution Images via Private Textual Intermediaries](https://mlsanity.com/paper/2506.07555/) (2025)
- [From Easy to Hard: Building a Shortcut for Differentially Private Image Synthesis](https://mlsanity.com/paper/2504.01395/) (2025)
- [Struct-Bench: A Benchmark for Differentially Private Structured Text Generation](https://mlsanity.com/paper/2509.10696/) (2025)
- [Private Evolution Converges](https://mlsanity.com/paper/2506.08312/) (2025)
- [PrivORL: Differentially Private Synthetic Dataset for Offline Reinforcement Learning](https://mlsanity.com/paper/2512.07342/) (2025)
- [Beyond One-Size-Fits-All: Neural Networks for Differentially Private Tabular Data Synthesis](https://mlsanity.com/paper/2511.13893/) (2025)
- [From Easy to Hard++: Promoting Differentially Private Image Synthesis Through Spatial-Frequency Curriculum](https://mlsanity.com/paper/2601.06368/) (2026)
- [ContinuousBench: Can Differentially Private Synthetic Text Improve Capabilities?](https://mlsanity.com/paper/2606.01849/) (2026)
- [Collaborative Synthetic Data Generation for Knowledge Transfer in Federated Learning](https://mlsanity.com/paper/2607.07565/) (2026)
- [Differentially Private Paired Table-Image Multimodal Synthesis](https://mlsanity.com/paper/2609.00708/) (2026)
- [QuanText: Protecting Dataset-Level Secrets in Textual Data Sharing](https://mlsanity.com/paper/2609.17995/) (2026)

## Abstract (arXiv)

Differentially private (DP) image synthesis aims to generate artificial images that retain the properties of sensitive images while protecting the privacy of individual images within the dataset. Despite recent advancements, we find that inconsistent--and sometimes flawed--evaluation protocols have been applied across studies. This not only impedes the understanding of current methods but also hinders future advancements. To address the issue, this paper introduces DPImageBench for DP image synthesis, with thoughtful design across several dimensions: (1) Methods. We study eleven prominent methods and systematically characterize each based on model architecture, pretraining strategy, and privacy mechanism. (2) Evaluation. We include nine datasets and seven fidelity and utility metrics to thoroughly assess them. Notably, we find that a common practice of selecting downstream classifiers based on the highest accuracy on the sensitive test set not only violates DP but also overestimates the utility scores. DPImageBench corrects for these mistakes. (3) Platform. Despite the methods and evaluation protocols, DPImageBench provides a standardized interface that accommodates current and future implementations within a unified framework. With DPImageBench, we have several noteworthy findings. For example, contrary to the common wisdom that pretraining on public image datasets is usually beneficial, we find that the distributional similarity between pretraining and sensitive images significantly impacts the performance of the synthetic images and does not always yield improvements. In addition, adding noise to low-dimensional features, such as the high-level characteristics of sensitive images, is less affected by the privacy budget compared to adding noise to high-dimensional features, like weight gradients. The former methods perform better than the latter under a low privacy budget.