PGD
A classical adversarial attack method that generates adversarial examples through iterative gradient ascent steps projected onto a perturbation constraint set. It is widely used to evaluate the robustness of machine learning models against adversarial inputs.
Timeline (last 90 days)
Systematic adversarial testing can use algorithms such as FGSM and PGD to generate adversarial samples, or employ Metamorphic Testing methods to construct semantically equivalent but formally different test inputs
Adversarial attack methods such as PGD and FGSM produce perturbations that are nearly imperceptible at the pixel level, yet can cause models to fail completely.
All Facts (2)
Systematic adversarial testing can use algorithms such as FGSM and PGD to generate adversarial samples, or employ Metamorphic Testing methods to construct semantically equivalent but formally different test inputs
50%UnverifiedAdversarial attack methods such as PGD and FGSM produce perturbations that are nearly imperceptible at the pixel level, yet can cause models to fail completely.
50%