Welcome to Adversary Detection Using Bayesian Approximations’s documentation!

Note

This is still a working project and documentation is in it’s skeleton form and therefore may be missing code coverage or code may be very loosely documented in places.

Computer Vision Models

For details and documentation surrounding computer vision based adversarial testing and Bayesian based classification:

Reinforcement Learning

Alternatively, a DQN has been trained to play FlappyBird with the effects of adversarial attacks upon the trained policy tested: