Evolutionary Adversarial Attacks on Deep Networks

Despite their uncontested success, recent studies have shown that Deep Neural Networks (DNNs) are vulnerable to adversarial attacks. A barely detectable change in an image can cause a misclassification in a well-trained DNN. Targeted adversarial examples can even evoke a misclassification of a specific class, for example, misclassifying a car as a cat. Researchers have demonstrated that adversarial attacks are successful in the real world and may be produced for data modalities beyond imaging, …

adversarial-attack adversarial attacks ai-generated image attacks car change craiyon deep learning deep-networks dnn evolutionary algorithms example examples generated genetic-algorithm image networks neural networks studies success vulnerable