We propose a novel setting for learning, where the input domain is the image of a map defined on the product of two sets, one of which completely determines the labels. We derive a new risk bound for this setting that decomposes into a bias and an error term, and exhibits a surprisingly weak dependence on the true labels. Inspired by these results, we present an algorithm aimed at minimizing the bias term by exploiting the ability to sample from each set independently. We apply our setting to visual classification tasks, where our approach enables us to train classifiers on datasets that consist entirely of a single synthetic example of each class. On several standard benchmarks for real-world image classification, we achieve robust performance in the context-agnostic setting, with good generalization to real world domains, whereas training directly on real world data without our techniques yields classifiers that are brittle to perturbations of the background.
翻译:我们提出一个新的学习环境, 输入域是两组产品中界定的地图的图像, 其中一组完全决定了标签。 我们为这个环境带来了一个新的风险, 它将分解成偏差和错误术语, 并展示出对真实标签的依赖性奇乎弱。 受这些结果的启发, 我们提出了一个算法, 目的是通过利用各组独立取样的能力, 最大限度地减少偏见。 我们将我们的设置应用到视觉分类任务中, 我们的方法使我们能够对分类人员进行完全由每类单一合成示例组成的数据集培训。 在现实世界图像分类的若干标准基准上, 我们取得了在背景不可知性环境中的强力表现, 向真实世界域作了很好的概括化, 而没有我们的技术, 直接进行真实世界数据的培训, 使得分类人员对背景的扭曲变得易碎乱。