Hardening Deep Neural Networks via Adversarial Model Cascades
2019 International Joint Conference on Neural Networks (IJCNN)(2018)
摘要
Deep neural networks (DNNs) are vulnerable to malicious inputs crafted by an adversary to produce erroneous outputs. Works on securing neural networks against adversarial examples achieve high empirical robustness on simple datasets such as MNIST. However, these techniques are inadequate when empirically tested on complex data sets such as CIFAR-10 and SVHN. Further, existing techniques are designed to target specific attacks and fail to generalize across attacks. We propose the Adversarial Model Cascades (AMC) as a way to tackle the above inadequacies. Our approach trains a cascade of models sequentially where each model is optimized to be robust towards a mixture of multiple attacks. Ultimately, it yields a single model which is secure against a wide range of attacks; namely FGSM, Elastic, Virtual Adversarial Perturbations and Madry. On an average, AMC increases the model's empirical robustness against various attacks simultaneously, by a significant margin (of 6.225 model's performance on non-adversarial inputs is comparable to the state-of-the-art models.
更多查看译文
关键词
deep neural networks,malicious inputs,erroneous outputs,adversarial examples,high empirical robustness,MNIST,complex data sets,CIFAR-10,multiple attacks,nonadversarial inputs,adversarial model cascades,AMC,SVHN,FGSM
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络