In this paper, we propose fuzzing-based attacks against machine learning models in hard-label black- box settings. We take the fuzz testing approach to generate ...
確定! 回上一頁