These are the codes used in the paper titled: "Heterogeneous Gaussian Mechanism: Preserving Differential Privacy in Deep Learning with Provable Robustness." (https://128.84.21.199/pdf/1906.01444.pdf) The 28th International Joint Conference on Artificial Intelligence (IJCAI 2019), August 10-16, 2019, Macao, China. (acceptance rate = 850/4752, 17.9%)
The software is written in tensorflow. It requires the following packages: python3, Tensorflow 1.1 or later.
If you use this code, please cite our paper.