Adversarial Training for Privacy-Preserving Deep Learning Model Distribution | IEEE Conference Publication | IEEE Xplore