DeepObliviate: A Powerful Charm for Erasing Data Residual Memory in Deep Neural Networks

no code implementations • 13 May 2021 • Yingzhe He, Guozhu Meng, Kai Chen, Jinwen He, Xingbo Hu

Compared to the method of retraining from scratch, our approach can achieve 99. 0%, 95. 0%, 91. 9%, 96. 7%, 74. 1% accuracy rates and 66. 7$\times$, 75. 0$\times$, 33. 3$\times$, 29. 4$\times$, 13. 7$\times$ speedups on the MNIST, SVHN, CIFAR-10, Purchase, and ImageNet datasets, respectively.

Machine Unlearning

Towards Security Threats of Deep Learning Systems: A Survey

no code implementations • 28 Nov 2019 • Yingzhe He, Guozhu Meng, Kai Chen, Xingbo Hu, Jinwen He

In order to unveil the security weaknesses and aid in the development of a robust deep learning system, we undertake an investigation on attacks towards deep learning, and analyze these attacks to conclude some findings in multiple views.

Adversarial Attack Model extraction