TranFuzz: AI Fuzzing and Black-Box Adversarial Attack Evaluation
Study and evaluation of TranFuzz for generating black-box adversarial examples on image classification models.
Cybersecurity
2 posts found
Study and evaluation of TranFuzz for generating black-box adversarial examples on image classification models.
Investigation and experimentation with laser fault injection techniques to analyze hardware vulnerabilities. Includes attack setups, power and pulse width parameters, and resulting BBICS behaviors.