Related papers: Mixed Differential Privacy in Computer Vision

Mixed Differential Privacy in Computer Vision

URL: http://arxiv.org/abs/2203.11481v1
Date: Tue, 22 Mar 2022 06:15:43 GMT
Title: Mixed Differential Privacy in Computer Vision
Authors: Aditya Golatkar, Alessandro Achille, Yu-Xiang Wang, Aaron Roth, Michael Kearns, Stefano Soatto
Abstract summary: AdaMix is an adaptive differentially private algorithm for training deep neural network classifiers using both private and public image data. A few-shot or even zero-shot learning baseline that ignores private data can outperform fine-tuning on a large private dataset.
Score: 133.68363478737058
License: http://creativecommons.org/licenses/by/4.0/
Abstract: We introduce AdaMix, an adaptive differentially private algorithm for training deep neural network classifiers using both private and public image data. While pre-training language models on large public datasets has enabled strong differential privacy (DP) guarantees with minor loss of accuracy, a similar practice yields punishing trade-offs in vision tasks. A few-shot or even zero-shot learning baseline that ignores private data can outperform fine-tuning on a large private dataset. AdaMix incorporates few-shot training, or cross-modal zero-shot learning, on public data prior to private fine-tuning, to improve the trade-off. AdaMix reduces the error increase from the non-private upper bound from the 167-311\% of the baseline, on average across 6 datasets, to 68-92\% depending on the desired privacy level selected by the user. AdaMix tackles the trade-off arising in visual classification, whereby the most privacy sensitive data, corresponding to isolated points in representation space, are also critical for high classification accuracy. In addition, AdaMix comes with strong theoretical privacy guarantees and convergence analysis.

Related papers

Unlocking Accuracy and Fairness in Differentially Private Image Classification [43.53494043189235]
Differential privacy (DP) is considered the gold standard framework for privacy-preserving training. We show that pre-trained foundation models fine-tuned with DP can achieve similar accuracy to non-private classifiers.
arXiv Detail & Related papers (2023-08-21T17:42:33Z)
Fine-Tuning with Differential Privacy Necessitates an Additional Hyperparameter Search [38.83524780461911]
We show how carefully selecting the layers being fine-tuned in the pretrained neural network allows us to establish new state-of-the-art tradeoffs between privacy and accuracy. We achieve 77.9% accuracy for $(varepsilon, delta)= (2, 10-5)$ on CIFAR-100 for a model pretrained on ImageNet.
arXiv Detail & Related papers (2022-10-05T11:32:49Z)
Individual Privacy Accounting for Differentially Private Stochastic Gradient Descent [69.14164921515949]
We characterize privacy guarantees for individual examples when releasing models trained by DP-SGD. We find that most examples enjoy stronger privacy guarantees than the worst-case bound. This implies groups that are underserved in terms of model utility simultaneously experience weaker privacy guarantees.
arXiv Detail & Related papers (2022-06-06T13:49:37Z)
Don't Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergence [73.14373832423156]
We propose DP-Sinkhorn, a novel optimal transport-based generative method for learning data distributions from private data with differential privacy. Unlike existing approaches for training differentially private generative models, we do not rely on adversarial objectives.
arXiv Detail & Related papers (2021-11-01T18:10:21Z)
On Deep Learning with Label Differential Privacy [54.45348348861426]
We study the multi-class classification setting where the labels are considered sensitive and ought to be protected. We propose a new algorithm for training deep neural networks with label differential privacy, and run evaluations on several datasets.
arXiv Detail & Related papers (2021-02-11T15:09:06Z)
Graph-Homomorphic Perturbations for Private Decentralized Learning [64.26238893241322]
Local exchange of estimates allows inference of data based on private data. perturbations chosen independently at every agent, resulting in a significant performance loss. We propose an alternative scheme, which constructs perturbations according to a particular nullspace condition, allowing them to be invisible.
arXiv Detail & Related papers (2020-10-23T10:35:35Z)
LDP-FL: Practical Private Aggregation in Federated Learning with Local Differential Privacy [20.95527613004989]
Federated learning is a popular approach for privacy protection that collects the local gradient information instead of real data. Previous works do not give a practical solution due to three issues. Last, the privacy budget explodes due to the high dimensionality of weights in deep learning models.
arXiv Detail & Related papers (2020-07-31T01:08:57Z)
Differentially private cross-silo federated learning [16.38610531397378]
Strict privacy is of paramount importance in distributed machine learning. In this paper we combine additively homomorphic secure summation protocols with differential privacy in the so-called cross-silo federated learning setting. We demonstrate that our proposed solutions give prediction accuracy that is comparable to the non-distributed setting.
arXiv Detail & Related papers (2020-07-10T18:15:10Z)
Differentially Private Federated Learning with Laplacian Smoothing [72.85272874099644]
Federated learning aims to protect data privacy by collaboratively learning a model without sharing private data among users. An adversary may still be able to infer the private training data by attacking the released model. Differential privacy provides a statistical protection against such attacks at the price of significantly degrading the accuracy or utility of the trained models.
arXiv Detail & Related papers (2020-05-01T04:28:38Z)

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