arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.11207cs.CVcs.LG
收藏

基于图像风格化的简单领域泛化

Frustratingly Simple Domain Generalization via Image Stylization

Georgia Institute of Technology(佐治亚理工学院)

Nathan Somavarapu, Chih-Yao Ma, Zsolt Kira

摘要:

卷积神经网络(CNN)在标准分类设置中表现出令人印象深刻的性能,其中训练和测试数据是独立同分布的。从给定的域。然而,CNN并不容易推广到具有不同统计数据的新领域,这对人类来说很简单。在这项工作中,我们解决了领域泛化问题,分类器必须推广到一个未知的目标域。受最近一些研究的启发,这些研究显示了CNN和人类之间的偏差差异,我们展示了一种非常简单但有效的方法,即通过用风格化图像增强数据集来纠正这种偏差。与现有的使用外部数据源(如艺术)的风格化作品相比,我们进一步引入了一种完全在域中使用的方法,不使用此类额外的数据源。我们提供了一个详细的分析,该方法的工作机制,验证我们的说法,它改变了形状/纹理的偏见,并证明结果超越或可比的最先进的,利用更复杂的方法。

英文摘要:

Convolutional Neural Networks (CNNs) show impressive performance in the standard classification setting where training and testing data are drawn i.i.d. from a given domain. However, CNNs do not readily generalize to new domains with different statistics, a setting that is simple for humans. In this work, we address the Domain Generalization problem, where the classifier must generalize to an unknown target domain. Inspired by recent works that have shown a difference in biases between CNNs and humans, we demonstrate an extremely simple yet effective method, namely correcting this bias by augmenting the dataset with stylized images. In contrast with existing stylization works, which use external data sources such as art, we further introduce a method that is entirely in-domain using no such extra sources of data. We provide a detailed analysis as to the mechanism by which the method works, verifying our claim that it changes the shape/texture bias, and demonstrate results surpassing or comparable to the state of the arts that utilize much more complex methods.

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。

↑