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 2101.09617cs.CV
收藏

一种深度模型鲁棒性综合评价框架

A Comprehensive Evaluation Framework for Deep Model Robustness

Beihang University(北京航空航天大学);China Electronics Standardization Institute(中国电子技术标准化研究院);Zhongguancun Laboratory(中关村实验室);Institute of Unmanned System, Beihang University(北京航空航天大学无人系统研究院);National Key Laboratory for Complex Systems Simulation(国家复杂系统仿真重点实验室)

Jun Guo, Wei Bao, Jiakai Wang, Yuqing Ma, Xinghai Gao, Gang Xiao, Aishan Liu, Jian Dong, Xianglong Liu, Wenjun Wu

摘要:

深度神经网络(DNN)在广泛的应用中取得了卓越的性能,但它们容易受到对抗性示例的影响,这激发了模型鲁棒性的评估和基准测试。然而,目前的评估通常使用简单的指标来研究防御的性能,这是远远没有了解这些防御方法的局限性和弱点。因此,大多数提出的防御很快就被证明是成功的攻击,这导致了攻击和防御之间的“军备竞赛”现象。为了缓解这个问题,我们建立了一个模型鲁棒性评估框架,其中包含23个全面而严格的指标,这些指标考虑了对抗学习的两个关键观点(即,数据和模型)。通过神经元覆盖和数据不可感知性,我们使用面向数据的度量来衡量测试示例的完整性;通过深入研究模型结构和行为,我们利用面向模型的度量来进一步评估对抗环境中的鲁棒性。为了充分证明我们框架的有效性,我们在CIFAR-10、SVHN和ImageNet等多个数据集上进行了大规模实验,使用我们的开源平台使用不同的模型和防御。总体而言,我们的论文提供了一个全面的评估框架,研究人员可以使用开源工具包进行全面和快速的评估,分析结果可以激发更深入的理解和进一步提高模型的鲁棒性。

英文摘要:

Deep neural networks (DNNs) have achieved remarkable performance across a wide range of applications, while they are vulnerable to adversarial examples, which motivates the evaluation and benchmark of model robustness. However, current evaluations usually use simple metrics to study the performance of defenses, which are far from understanding the limitation and weaknesses of these defense methods. Thus, most proposed defenses are quickly shown to be attacked successfully, which results in the ``arm race'' phenomenon between attack and defense. To mitigate this problem, we establish a model robustness evaluation framework containing 23 comprehensive and rigorous metrics, which consider two key perspectives of adversarial learning (i.e., data and model). Through neuron coverage and data imperceptibility, we use data-oriented metrics to measure the integrity of test examples; by delving into model structure and behavior, we exploit model-oriented metrics to further evaluate robustness in the adversarial setting. To fully demonstrate the effectiveness of our framework, we conduct large-scale experiments on multiple datasets including CIFAR-10, SVHN, and ImageNet using different models and defenses with our open-source platform. Overall, our paper provides a comprehensive evaluation framework, where researchers could conduct comprehensive and fast evaluations using the open-source toolkit, and the analytical results could inspire deeper understanding and further improvement to the model robustness.

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。

↑