Image-mixing augmentations (e.g., Mixup or CutMix), which typically mix two images, have become de-facto training tricks for image classification. Despite their huge success on image classification, the number of images to mix has not been profoundly investigated by the previous works, only showing the naive K-image expansion leads to poor performance degradation. This paper derives a new K-image mixing augmentation based on the stick-breaking process under Dirichlet prior. We show that our method can train more robust and generalized classifiers through extensive experiments and analysis on classification accuracy, a shape of a loss landscape and adversarial robustness, than the usual two-image methods. Furthermore, we show that our probabilistic model can measure the sample-wise uncertainty and can boost the efficiency for Network Architecture Search (NAS) with 7x reduced search time.
翻译:图像混合放大器(例如,Mixup 或 CutMix ) 通常混合两种图像,它们已成为图像分类的不折不扣的培训技巧。 尽管在图像分类上取得了巨大成功,但先前的作品并没有深入地调查混合图像的数量,只是展示了天真的K图像扩展导致性能退化。本文根据Drichlet 之前的刺破过程产生了一种新的K图像混合放大器。我们显示,我们的方法可以通过对分类准确性、损失景观形状和对抗性强力进行广泛的实验和分析,对分类精度、损失景观和对抗性强力进行更加稳健和普及的训练。 此外,我们显示,我们的概率模型可以测量样本不确定性,并能提高网络架构搜索的效率,减少7x搜索时间。