On instabilities of deep learning in image reconstruction and the potential costs of AI
On instabilities of deep learning in image reconstruction and the potential costs of AI
复制标题
DOI：
10.1073/pnas.1907377117
复制
发表时间：
2020-12-01
影响因子：
11.1
通讯作者：
Hansen, Anders C.
中科院分区：
文献类型：
--
作者：
Antun, Vegard;Renna, Francesco;Hansen, Anders C.
由于深度学习在图像分类等任务中取得了前所未有的成功，它已成为图像重建的新工具，有可能改变该领域。在本文中，我们展示了一个关键现象：深度学习通常会产生不稳定的图像重建方法。不稳定性通常以几种形式出现：1）在图像和采样域中的某些微小的、几乎不可检测的扰动可能导致重建中的严重伪影; 2）小的结构变化，例如肿瘤，可能无法在重建图像中捕获;以及3）（违反直觉类型的不稳定性）更多的样本可能产生更差的性能。我们的稳定性测试使用算法和易于使用的软件检测不稳定现象。该测试针对的是研究人员，以测试他们的网络是否不稳定，以及美国食品和药物管理局（FDA）等政府机构，以确保深度学习方法的安全使用。
Deep learning, due to its unprecedented success in tasks such as image classification, has emerged as a new tool in image reconstruction with potential to change the field. In this paper, we demonstrate a crucial phenomenon: Deep learning typically yields unstable methods for image reconstruction. The instabilities usually occur in several forms: 1) Certain tiny, almost undetectable perturbations, both in the image and sampling domain, may result in severe artefacts in the reconstruction; 2) a small structural change, for example, a tumor, may not be captured in the reconstructed image; and 3) (a counterintuitive type of instability) more samples may yield poorer performance. Our stability test with algorithms and easy-to-use software detects the instability phenomena. The test is aimed at researchers, to test their networks for instabilities, and for government agencies, such as the Food and Drug Administration (FDA), to secure safe use of deep learning methods.