Detecting Operational Adversarial Examples for Reliable Deep Learning | IEEE Conference Publication | IEEE Xplore