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 2003.03879cs.CV
收藏

正则化方法稳健性和不确定性的实证评价

An Empirical Evaluation on Robustness and Uncertainty of Regularization Methods

Clova AI Research, NAVER Corp.(克洛娃人工智能研究院,NAVER公司);Clova AI Research, LINE Plus Corp.(克洛娃人工智能研究院,LINE Plus公司);Yonsei University(延世大学)

Sanghyuk Chun, Seong Joon Oh, Sangdoo Yun, Dongyoon Han, Junsuk Choe, Youngjoon Yoo

摘要:

尽管深度神经网络(DNN)具有明显的人类水平的性能,但它们的行为与人类有着根本的不同。当模糊和噪声等小的损坏应用于输入时,它们很容易改变预测(缺乏鲁棒性),并且它们经常对分布外的样本产生置信预测(不正确的不确定性度量)。虽然许多研究旨在解决这些问题,但提出的解决方案通常昂贵且复杂(例如贝叶斯推理和对抗训练)。同时,许多简单而廉价的正则化方法也被开发出来以提高分类器的泛化能力。这种正则化方法在很大程度上被忽视了作为解决鲁棒性和不确定性问题的基线,因为它们不是专门为此设计的。在本文中,我们对使用最先进的正则化方法训练的图像分类器(CIFAR-100和ImageNet)的鲁棒性和不确定性估计进行了广泛的实证评估。此外,实验结果表明,某些正则化方法可以作为DNN的鲁棒性和不确定性估计的强基线方法。

英文摘要:

Despite apparent human-level performances of deep neural networks (DNN), they behave fundamentally differently from humans. They easily change predictions when small corruptions such as blur and noise are applied on the input (lack of robustness), and they often produce confident predictions on out-of-distribution samples (improper uncertainty measure). While a number of researches have aimed to address those issues, proposed solutions are typically expensive and complicated (e.g. Bayesian inference and adversarial training). Meanwhile, many simple and cheap regularization methods have been developed to enhance the generalization of classifiers. Such regularization methods have largely been overlooked as baselines for addressing the robustness and uncertainty issues, as they are not specifically designed for that. In this paper, we provide extensive empirical evaluations on the robustness and uncertainty estimates of image classifiers (CIFAR-100 and ImageNet) trained with state-of-the-art regularization methods. Furthermore, experimental results show that certain regularization methods can serve as strong baseline methods for robustness and uncertainty estimation of DNNs.

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。

↑