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 2003.10602cs.LGstat.ML
收藏

通过不同方向进行防御

Defense Through Diverse Directions

The University of North Carolina(北卡罗来纳大学)

Christopher M. Bender, Yang Li, Yifeng Shi, Michael K. Reiter, Junier B. Oliva

摘要:

在这项工作中,我们开发了一种新的贝叶斯神经网络方法,以实现强大的对抗鲁棒性,而不需要在线对抗训练。与以前在这个方向上的努力不同,我们不依赖于网络权重的随机性,通过最小化学习参数分布和先验之间的分歧。相反,我们还要求模型对所有输入协变量保持一定的预期不确定性。我们证明,通过鼓励网络在输入之间均匀分布,网络变得不太容易受到局部化的,脆弱的功能,赋予一个自然的鲁棒性,有针对性的扰动。我们在几个基准数据集上显示了经验鲁棒性。

英文摘要:

In this work we develop a novel Bayesian neural network methodology to achieve strong adversarial robustness without the need for online adversarial training. Unlike previous efforts in this direction, we do not rely solely on the stochasticity of network weights by minimizing the divergence between the learned parameter distribution and a prior. Instead, we additionally require that the model maintain some expected uncertainty with respect to all input covariates. We demonstrate that by encouraging the network to distribute evenly across inputs, the network becomes less susceptible to localized, brittle features which imparts a natural robustness to targeted perturbations. We show empirical robustness on several benchmark datasets.

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。

↑