Developing and Defeating Adversarial Examples
This work addresses safety concerns for systems using DNNs in critical applications, but it is incremental as it applies known adversarial techniques to a specific model.
The paper tackled the problem of adversarial examples in deep neural networks by developing attacks on the Yolo V3 object detector and studying detection and neutralization strategies, resulting in available Python code for implementation.
Breakthroughs in machine learning have resulted in state-of-the-art deep neural networks (DNNs) performing classification tasks in safety-critical applications. Recent research has demonstrated that DNNs can be attacked through adversarial examples, which are small perturbations to input data that cause the DNN to misclassify objects. The proliferation of DNNs raises important safety concerns about designing systems that are robust to adversarial examples. In this work we develop adversarial examples to attack the Yolo V3 object detector [1] and then study strategies to detect and neutralize these examples. Python code for this project is available at https://github.com/ianmcdiarmidsterling/adversarial