arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.04027cs.LGcs.CVcs.RO
收藏

基于混合最大熵变换的填充级分类数据增强

Data augmentation with mixtures of max-entropy transformations for filling-level classification

EPFL(洛桑联邦理工学院);Queen Mary University of London(伦敦玛丽女王大学)

Apostolos Modas, Andrea Cavallaro, Pascal Frossard

摘要:

我们解决的问题,在测试时的数据与原则的数据增强计划的任务,内容级分类的分布偏移。在这样的任务中,诸如测试时容器(杯子或饮用玻璃杯)的形状或透明度之类的属性可能不同于训练数据中表示的那些属性。使用标准增强方案处理这种分布偏移是具有挑战性的,并且变换训练图像以覆盖测试时间实例的属性需要复杂的图像操作。因此,我们使用一系列最大熵变换来生成不同的增强,这些最大熵变换可以创建具有新形状、颜色和光谱特性的样本。我们表明,这样一个原则性的增强方案,单独,可以取代目前的方法,使用迁移学习或可以与迁移学习结合使用,以提高其性能。

英文摘要:

We address the problem of distribution shifts in test-time data with a principled data augmentation scheme for the task of content-level classification. In such a task, properties such as shape or transparency of test-time containers (cup or drinking glass) may differ from those represented in the training data. Dealing with such distribution shifts using standard augmentation schemes is challenging and transforming the training images to cover the properties of the test-time instances requires sophisticated image manipulations. We therefore generate diverse augmentations using a family of max-entropy transformations that create samples with new shapes, colors and spectral characteristics. We show that such a principled augmentation scheme, alone, can replace current approaches that use transfer learning or can be used in combination with transfer learning to improve its performance.

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。

↑