T. Wang, Y. Zhang, and R. Jia, “Improving Robustness to Model Inversion Attacks via Mutual Information Regularization”, AAAI, vol. 35, no. 13, pp. 11666-11673, May 2021.