Tag: Baidu

Baidu details its adversarial toolbox for testing robustness of AI models

No matter the claimed robustness of AI and machine learning systems in production, none are immune to adversarial attacks, or techniques that attempt to fool algorithms through malicious input. It’s been shown that generating even small perturbations on images can fool the best of classifiers with high probability. And that’s problematic considering the wide proliferation […]

Read More