Reinforcing the Robustness of a Deep Neural Network to Adversarial Examples by Using Color Quantization of Training Image Data | IEEE Conference Publication | IEEE Xplore