1 paper
Yang Zheng, Xiaoyi Feng, Zhaoqiang Xia +5
Adversarial reprogramming allows stealing computational resources by repurposing machine learning models to perform a different task chosen by the attacker. For example, a model tr…