1 paper · 1 filter
Frank Mollard, Marcus Becker, Florian Roehrbein
The paper introduces a white-box attack on computer vision models using SHAP values. It demonstrates how adversarial evasion attacks can compromise the performance of deep learning…