arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.08978cs.CVcs.AIcs.LG
收藏

周期性病灶丢失

Cyclical Focal Loss

U.S. Naval Research Laboratory(美国海军研究实验室);Naval Center for Applied Research in AI(海军人工智能应用研究中心)

Leslie N. Smith

摘要:

交叉熵softmax损失是用于训练深度神经网络的主要损失函数。另一方面,焦点损失函数已被证明在每个类中的训练样本的数量不平衡时提供改进的性能,例如在长尾数据集中。在本文中,我们介绍了一种新的循环焦点损失,并证明了它是一个更普遍的损失函数比交叉熵softmax损失或焦点损失。我们描述了周期性焦点丢失背后的直觉,我们的实验提供了证据,证明周期性焦点丢失为平衡,不平衡或长尾数据集提供了卓越的性能。我们提供了CIFAR-10/CIFAR-100,ImageNet,CIFAR-10/CIFAR-100的平衡和不平衡4,000个训练样本版本以及开放长尾识别(OLTR)挑战中的ImageNet-LT和Places-LT的大量实验结果。实现循环焦点损失函数只需要几行代码,并且不会增加训练时间。本着可重复性的精神,我们的代码可在\url{https://github.com/lnsmith54/CFL}上获得。

英文摘要:

The cross-entropy softmax loss is the primary loss function used to train deep neural networks. On the other hand, the focal loss function has been demonstrated to provide improved performance when there is an imbalance in the number of training samples in each class, such as in long-tailed datasets. In this paper, we introduce a novel cyclical focal loss and demonstrate that it is a more universal loss function than cross-entropy softmax loss or focal loss. We describe the intuition behind the cyclical focal loss and our experiments provide evidence that cyclical focal loss provides superior performance for balanced, imbalanced, or long-tailed datasets. We provide numerous experimental results for CIFAR-10/CIFAR-100, ImageNet, balanced and imbalanced 4,000 training sample versions of CIFAR-10/CIFAR-100, and ImageNet-LT and Places-LT from the Open Long-Tailed Recognition (OLTR) challenge. Implementing the cyclical focal loss function requires only a few lines of code and does not increase training time. In the spirit of reproducibility, our code is available at \url{https://github.com/lnsmith54/CFL}.

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。

↑