VoronoiPatches: Evaluating A New Data Augmentation Method

TL;DR

This paper introduces VoronoiPatches, a novel data augmentation technique using convex polygon patches to improve CNN generalization by reducing overfitting and increasing robustness on unseen data.

Contribution

The paper presents VoronoiPatches, a new data augmentation method that employs non-linear recombination with polygon patches, outperforming existing methods in reducing overfitting.

Findings

VoronoiPatches outperforms existing data augmentation methods in reducing model variance.

VP enhances CNN robustness on unseen data.

Smoothing optional transitions improves augmentation effectiveness.

Abstract

Overfitting is a problem in Convolutional Neural Networks (CNN) that causes poor generalization of models on unseen data. To remediate this problem, many new and diverse data augmentation methods (DA) have been proposed to supplement or generate more training data, and thereby increase its quality. In this work, we propose a new data augmentation algorithm: VoronoiPatches (VP). We primarily utilize non-linear recombination of information within an image, fragmenting and occluding small information patches. Unlike other DA methods, VP uses small convex polygon-shaped patches in a random layout to transport information around within an image. Sudden transitions created between patches and the original image can, optionally, be smoothed. In our experiments, VP outperformed current DA methods regarding model variance and overfitting tendencies. We demonstrate data augmentation utilizing non-linear re-combination of information within images, and non-orthogonal shapes and structures improves CNN model robustness on unseen data.