arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2004.10162cs.LGcs.CVstat.ML
收藏

EMPIR:混合精度深度网络的集合,用于增强对对抗攻击的鲁棒性

EMPIR: Ensembles of Mixed Precision Deep Networks for Increased Robustness against Adversarial Attacks

Purdue University(普渡大学);Indian Institute of Technology (IIT) Madras(印度理工学院马德拉斯分校)

Sanchari Sen, Balaraman Ravindran, Anand Raghunathan

摘要:

确保深度神经网络(DNN)的鲁棒性对于它们在自动驾驶汽车、无人机和医疗保健等安全关键型应用中的采用至关重要。值得注意的是,DNN容易受到对抗攻击的攻击,其中小的输入扰动可能会产生灾难性的错误分类。在这项工作中,我们提出了EMPIR,具有不同数值精度的量化DNN模型的集合,作为一种新的方法来提高对对抗攻击的鲁棒性。EMPIR是基于这样的观察,即量化神经网络通常比全精度网络对对抗攻击表现出更高的鲁棒性,但代价是原始(未扰动)输入的准确性大幅下降。EMPIR克服了这一限制,实现了“两全其美”,即,全精度模型的更高的无扰动精度与低精度模型的更高的鲁棒性相结合,通过将它们组成一个集合。此外,由于低精度DNN模型的计算和存储需求比全精度模型低得多,因此与单个全精度模型相比,EMPIR模型只会产生适度的计算和内存开销(在我们的评估中<25%)。我们在一套DNN上评估EMPIR,用于3种不同的图像识别任务(MNIST,CIFAR-10和ImageNet)和4种不同的对抗攻击。我们的研究结果表明,与单个全精度模型相比,EMPIR分别将MNIST,CIFAR-10和ImageNet数据集上训练的DNN模型的平均对抗准确率提高了42.6%,15.2%和10.5%,而不会牺牲未扰动输入的准确性。

英文摘要:

Ensuring robustness of Deep Neural Networks (DNNs) is crucial to their adoption in safety-critical applications such as self-driving cars, drones, and healthcare. Notably, DNNs are vulnerable to adversarial attacks in which small input perturbations can produce catastrophic misclassifications. In this work, we propose EMPIR, ensembles of quantized DNN models with different numerical precisions, as a new approach to increase robustness against adversarial attacks. EMPIR is based on the observation that quantized neural networks often demonstrate much higher robustness to adversarial attacks than full precision networks, but at the cost of a substantial loss in accuracy on the original (unperturbed) inputs. EMPIR overcomes this limitation to achieve the 'best of both worlds', i.e., the higher unperturbed accuracies of the full precision models combined with the higher robustness of the low precision models, by composing them in an ensemble. Further, as low precision DNN models have significantly lower computational and storage requirements than full precision models, EMPIR models only incur modest compute and memory overheads compared to a single full-precision model (<25% in our evaluations). We evaluate EMPIR across a suite of DNNs for 3 different image recognition tasks (MNIST, CIFAR-10 and ImageNet) and under 4 different adversarial attacks. Our results indicate that EMPIR boosts the average adversarial accuracies by 42.6%, 15.2% and 10.5% for the DNN models trained on the MNIST, CIFAR-10 and ImageNet datasets respectively, when compared to single full-precision models, without sacrificing accuracy on the unperturbed inputs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑