Interesting #ml #security paper at #ICCV2021: "data-free #adversarial perturbations for practical black-box #attack" I've created a summary on a security #flashcard and added it to security-flashcards.com @ICCV_2021