Fetching the paper…

Defending Model Inversion and Membership Inference Attacks via Prediction Purification · Around