Code Resources from a Rhodes Artificial Intelligence Lab presentation on AI and Cybersecurity
All of the included files are intended to run on Google's colab environment.
RAIL_Captcha_Learning - this file is where we trained the captcha cracking neural network
RAIL_Captcha_Attack - this file is where we used our trained model to crack actual captchas
RAIL_Adversarial_Concealment - this file has examples of adversarials which are not targeted. We used a bridge as our input image.
RAIL_Adversarial_Deception - this file has examples of targeted adversarials. We used a kitten as our input image.