1 paper
Zhen Liang, Taoran Wu, Ran Zhao +5
Recently, formal verification of deep neural networks (DNNs) has garnered considerable attention, and over-approximation based methods have become popular due to their effectivenes…