论文标题
一种基于GAN的图像转换方案，用于保护隐私的深度神经网络
A GAN-Based Image Transformation Scheme for Privacy-Preserving Deep Neural Networks
论文作者
论文摘要
我们建议使用生成对抗网络（GAN）提出一种新型的图像转化方案，以保护隐私的深度神经网络（DNNS）。提出的方案不仅使我们不仅可以在没有视觉信息的情况下将图像应用于DNN，还可以增强针对仅密文的仅限攻击（COA）（包括基于DNN的攻击）的鲁棒性。在本文中，提出的转换方案被证明能够保护纯图像上的视觉信息，并且视觉保护的图像直接应用于DNNS以进行隐私图像分类。由于所提出的方案利用了GAN，因此无需管理加密密钥。在图像分类实验中，我们根据针对COA的分类准确性和鲁棒性评估了所提出的方案的有效性。
We propose a novel image transformation scheme using generative adversarial networks (GANs) for privacy-preserving deep neural networks (DNNs). The proposed scheme enables us not only to apply images without visual information to DNNs, but also to enhance robustness against ciphertext-only attacks (COAs) including DNN-based attacks. In this paper, the proposed transformation scheme is demonstrated to be able to protect visual information on plain images, and the visually-protected images are directly applied to DNNs for privacy-preserving image classification. Since the proposed scheme utilizes GANs, there is no need to manage encryption keys. In an image classification experiment, we evaluate the effectiveness of the proposed scheme in terms of classification accuracy and robustness against COAs.
