arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.15475cs.CV
收藏

异构标签噪声如何影响神经网络中的泛化?

How Does Heterogeneous Label Noise Impact Generalization in Neural Nets?

Rochester Institute of Technology(罗切斯特理工学院);Paige(佩奇公司);Cornell Tech(康奈尔科技学院)

Bidur Khanal, Christopher Kanan

摘要:

不正确标记的示例或标签噪声在现实世界的计算机视觉数据集中很常见。虽然在先前的工作中已经研究了标签噪声对深度神经网络中的学习的影响,但是这些研究仅仅集中在同质标签噪声上,即,标签噪声的程度在所有类别中是相同的。然而,在现实世界中,标签噪声往往是异构的,某些类别受到的影响程度比其他类别更大。在这里,我们解决了文献中的这一差距。我们假设异质标签噪声只会影响有标签噪声的类,除非从这些类转移到没有标签噪声的类。为了验证这一假设,我们设计了一系列使用MNIST、CIFAR-10、CIFAR-100和MS-COCO的计算机视觉研究,其中我们在多类、多任务和多标签系统的训练过程中施加了异质标签噪声。我们的研究结果提供了证据支持我们的假设:标签噪声只影响受其影响的类,除非有转移。

英文摘要:

Incorrectly labeled examples, or label noise, is common in real-world computer vision datasets. While the impact of label noise on learning in deep neural networks has been studied in prior work, these studies have exclusively focused on homogeneous label noise, i.e., the degree of label noise is the same across all categories. However, in the real-world, label noise is often heterogeneous, with some categories being affected to a greater extent than others. Here, we address this gap in the literature. We hypothesized that heterogeneous label noise would only affect the classes that had label noise unless there was transfer from those classes to the classes without label noise. To test this hypothesis, we designed a series of computer vision studies using MNIST, CIFAR-10, CIFAR-100, and MS-COCO where we imposed heterogeneous label noise during the training of multi-class, multi-task, and multi-label systems. Our results provide evidence in support of our hypothesis: label noise only affects the class affected by it unless there is transfer.

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。

↑