RandoMix: A mixed sample data augmentation method with multiple mixed modes

TL;DR

RandoMix is a versatile data augmentation technique combining linear and mask-mixed modes, significantly improving model robustness and diversity across multiple datasets and outperforming existing methods.

Contribution

It introduces RandoMix, a novel mixed-sample augmentation method that enhances robustness and diversity by combining multiple mixing modes with flexible candidate and weight selection.

Findings

Outperforms Mixup, CutMix, Fmix, and ResizeMix in accuracy.

Enhances robustness against adversarial and natural noise.

Effective across diverse datasets including CIFAR, ImageNet, and speech commands.

Abstract

Data augmentation plays a crucial role in enhancing the robustness and performance of machine learning models across various domains. In this study, we introduce a novel mixed-sample data augmentation method called RandoMix. RandoMix is specifically designed to simultaneously address robustness and diversity challenges. It leverages a combination of linear and mask-mixed modes, introducing flexibility in candidate selection and weight adjustments. We evaluate the effectiveness of RandoMix on diverse datasets, including CIFAR-10/100, Tiny-ImageNet, ImageNet, and Google Speech Commands. Our results demonstrate its superior performance compared to existing techniques such as Mixup, CutMix, Fmix, and ResizeMix. Notably, RandoMix excels in enhancing model robustness against adversarial noise, natural noise, and sample occlusion. The comprehensive experimental results and insights into parameter tuning underscore the potential of RandoMix as a versatile and effective data augmentation method. Moreover, it seamlessly integrates into the training pipeline.