Deep neural networks have become popular in many supervised learning tasks, but they may suffer from overfitting when the training dataset is limited. To mitigate this, many researchers use data augmentation, which is a widely used and effective method for increasing the variety of datasets. However, the randomness introduced by data augmentation causes inevitable inconsistency between training and inference, which leads to poor improvement. In this paper, we propose a consistency regularization framework based on data augmentation, called CR-Aug, which forces the output distributions of different sub models generated by data augmentation to be consistent with each other. Specifically, CR-Aug evaluates the discrepancy between the output distributions of two augmented versions of each sample, and it utilizes a stop-gradient operation to minimize the consistency loss. We implement CR-Aug to image and audio classification tasks and conduct extensive experiments to verify its effectiveness in improving the generalization ability of classifiers. Our CR-Aug framework is ready-to-use, it can be easily adapted to many state-of-the-art network architectures. Our empirical results show that CR-Aug outperforms baseline methods by a significant margin.
翻译:深神经网络在许多受监督的学习任务中变得很受欢迎,但当培训数据集有限时,深神经网络可能因过度配置而受到影响。为了缓解这一点,许多研究人员使用数据增强,这是广泛使用的有效方法,可以增加数据集的多样性。然而,数据增强带来的随机性不可避免地造成培训和推论之间的不一致,导致改进不力。在本文件中,我们提出了一个基于数据增强的一致规范化框架,称为CR-Aug,它迫使数据增强产生的不同子模型的输出分布相互一致。具体地说,CR-Aug评估了每个样本两个扩大版本的输出分布之间的差异,并使用了截分级操作,以尽量减少一致性损失。我们实施了CR-Aug,以图像和音频分类任务,并进行了广泛的实验,以核实其在提高分类员普遍化能力方面的有效性。我们的CR-Aug框架已经准备使用,很容易适应许多州级网络结构。我们的经验结果表明,CR-Aug在显著的幅度上超越了基线方法。