No Free Lunch for Synthetic Images under Data Scarcity Conditions

• URL: http://arxiv.org/abs/2606.07640v1
• Date: Mon, 01 Jun 2026 12:45:02 GMT
• Title: No Free Lunch for Synthetic Images under Data Scarcity Conditions
• Abstract summary: This study investigates the trade-offs between fidelity, privacy, and utility in synthetic data generation under conditions of data scarcity and privacy sensitivity.<n>We propose an evaluation framework that jointly assesses these three dimensions and apply it to three widely used generative models, VAE, GAN, and DDPM.<n>The evaluation spans three image datasets, MNIST, OCTMNIST, and OrganAMNIST, encompassing both general-purpose and medical imaging domains.
• Score: 38.510551031367584
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: This study investigates the trade-offs between fidelity, privacy, and utility in synthetic data generation under conditions of data scarcity and privacy sensitivity. We propose an evaluation framework that jointly assesses these three dimensions and apply it to three widely used generative models, VAE, GAN, and DDPM. The evaluation spans three image datasets, MNIST, OCTMNIST, and OrganAMNIST, encompassing both general-purpose and medical imaging domains. Notable differences arise between the three models in their behaviour when differential privacy mechanisms are introduced during training. GAN and DDPM demonstrate greater robustness, maintaining higher fidelity and downstream utility across a range of noise levels, while VAE degrades more rapidly as privacy constraints increase. This study highlights the importance of a multidimensional evaluation of deep generative models, also noting that their behaviour significantly differs when privacy techniques are applied.

Related papers

• Class-Aware Adaptive Differential Privacy in Deep Learning for Sensor-Based Fall Detection [0.0]
  Fall detection is a critical task in healthcare, particularly for elderly people.<n>Previous privacy approaches apply uniform noise across all training samples, which affects the prediction performance.<n>We propose a Class-Aware Adaptive Differential Privacy framework integrated with a hybrid 3D Convolutional Neural Network and Bidirectional Long Short-Term Memory architecture.
  arXiv  Detail & Related papers  (2026-05-03T02:34:54Z)
• Balancing Fidelity, Utility, and Privacy in Synthetic Cardiac MRI Generation: A Comparative Study [0.43217519111088665]
  Deep learning in cardiac MRI (CMR) is fundamentally constrained by both data scarcity and privacy regulations.<n>This study benchmarks three generative architectures: Denoising Diffusion Probabilistic Models (DDPM), Latent Diffusion Models (LDM), and Flow Matching (FM)<n>We evaluate generated data across three critical axes: fidelity, utility, and privacy.
  arXiv  Detail & Related papers  (2026-03-04T17:59:00Z)
• FusionDP: Foundation Model-Assisted Differentially Private Learning for Partially Sensitive Features [17.945111987608865]
  In practical scenarios, privacy protection may be required for only a subset of features.<n>Traditional DP-SGD enforces privacy protection on all features in one sample, leading to excessive noise injection and significant utility degradation.<n>We propose FusionDP, a two-step framework that enhances model utility under feature-level differential privacy.
  arXiv  Detail & Related papers  (2025-11-05T19:13:10Z)
• On the MIA Vulnerability Gap Between Private GANs and Diffusion Models [51.53790101362898]
  Generative Adversarial Networks (GANs) and diffusion models have emerged as leading approaches for high-quality image synthesis.<n>We present the first unified theoretical and empirical analysis of the privacy risks faced by differentially private generative models.
  arXiv  Detail & Related papers  (2025-09-03T14:18:22Z)
• On Differentially Private 3D Medical Image Synthesis with Controllable Latent Diffusion Models [5.966954237899151]
  This study addresses challenges for 3D cardiac MRI images in the short-axis view. We propose Latent Diffusion Models that generate synthetic images conditioned on medical attributes. We finetune our models with differential privacy on the UK Biobank dataset.
  arXiv  Detail & Related papers  (2024-07-23T11:49:58Z)
• PATE-TripleGAN: Privacy-Preserving Image Synthesis with Gaussian Differential Privacy [4.586288671392977]
  We present a privacy-preserving training framework called PATE-TripleGAN. It incorporates a classifier to pre-classify unlabeled data to reduce dependence on labeled data. PATE-TripleGAN can generate a higher quality labeled image dataset while ensuring privacy of the training data.
  arXiv  Detail & Related papers  (2024-04-19T09:22:20Z)
• TernaryVote: Differentially Private, Communication Efficient, and Byzantine Resilient Distributed Optimization on Heterogeneous Data [50.797729676285876]
  We propose TernaryVote, which combines a ternary compressor and the majority vote mechanism to realize differential privacy, gradient compression, and Byzantine resilience simultaneously. We theoretically quantify the privacy guarantee through the lens of the emerging f-differential privacy (DP) and the Byzantine resilience of the proposed algorithm.
  arXiv  Detail & Related papers  (2024-02-16T16:41:14Z)
• Shielding the Unseen: Privacy Protection through Poisoning NeRF with Spatial Deformation [59.302770084115814]
  We introduce an innovative method of safeguarding user privacy against the generative capabilities of Neural Radiance Fields (NeRF) models. Our novel poisoning attack method induces changes to observed views that are imperceptible to the human eye, yet potent enough to disrupt NeRF's ability to accurately reconstruct a 3D scene. We extensively test our approach on two common NeRF benchmark datasets consisting of 29 real-world scenes with high-quality images.
  arXiv  Detail & Related papers  (2023-10-04T19:35:56Z)
• Private, fair and accurate: Training large-scale, privacy-preserving AI models in medical imaging [47.99192239793597]
  We evaluated the effect of privacy-preserving training of AI models regarding accuracy and fairness compared to non-private training. Our study shows that -- under the challenging realistic circumstances of a real-life clinical dataset -- the privacy-preserving training of diagnostic deep learning models is possible with excellent diagnostic accuracy and fairness.
  arXiv  Detail & Related papers  (2023-02-03T09:49:13Z)
• Robustness Threats of Differential Privacy [70.818129585404]
  We experimentally demonstrate that networks, trained with differential privacy, in some settings might be even more vulnerable in comparison to non-private versions. We study how the main ingredients of differentially private neural networks training, such as gradient clipping and noise addition, affect the robustness of the model.
  arXiv  Detail & Related papers  (2020-12-14T18:59:24Z)
• Hide-and-Seek Privacy Challenge [88.49671206936259]
  The NeurIPS 2020 Hide-and-Seek Privacy Challenge is a novel two-tracked competition to accelerate progress in tackling both problems. In our head-to-head format, participants in the synthetic data generation track (i.e. "hiders") and the patient re-identification track (i.e. "seekers") are directly pitted against each other by way of a new, high-quality intensive care time-series dataset.
  arXiv  Detail & Related papers  (2020-07-23T15:50:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.