论文标题
部分可观测时空混沌系统的无模型预测
Attention Consistency on Visual Corruptions for Single-Source Domain Generalization
论文作者
论文摘要
概括了对单个分布进行训练的视觉识别模型（即域），这需要使它们与训练集中的多余相关性保持稳健。在这项工作中，我们通过更改训练图像来模拟新域并在同一样本的不同视图中引起一致的视觉关注来实现这一目标。我们发现，第一个目标可以通过视觉腐败简单有效地实现。具体而言，我们使用Imagenet-C基准的19个损坏和基于傅立叶变换的其他三个转换来更改训练图像的内容。由于这些腐败保护对象位置，因此我们提出了注意力一致性损失，以确保对同一训练样本的原始和损坏版本的班级激活图保持一致。我们将视觉损坏（ACVC）的模型关注一致性命名。我们表明，ACVC始终在三个单源域概括基准，PACS，可可和大规模域上实现最新技术。
Generalizing visual recognition models trained on a single distribution to unseen input distributions (i.e. domains) requires making them robust to superfluous correlations in the training set. In this work, we achieve this goal by altering the training images to simulate new domains and imposing consistent visual attention across the different views of the same sample. We discover that the first objective can be simply and effectively met through visual corruptions. Specifically, we alter the content of the training images using the nineteen corruptions of the ImageNet-C benchmark and three additional transformations based on Fourier transform. Since these corruptions preserve object locations, we propose an attention consistency loss to ensure that class activation maps across original and corrupted versions of the same training sample are aligned. We name our model Attention Consistency on Visual Corruptions (ACVC). We show that ACVC consistently achieves the state of the art on three single-source domain generalization benchmarks, PACS, COCO, and the large-scale DomainNet.
