Feature transforms for image data augmentation

TL;DR

This paper introduces new image augmentation techniques based on Fourier, Radon, and Discrete Cosine Transforms, and demonstrates that combining these with existing methods enhances CNN robustness on small datasets.

Contribution

The work proposes three novel data augmentation methods using FT, RT, and DCT transforms and shows their effectiveness in ensemble models for image classification.

Findings

Ensemble methods with new augmentations outperform existing techniques.

Transform-based augmentations improve CNN robustness on small datasets.

Combining multiple augmentation approaches yields state-of-the-art results.

Abstract

A problem with Convolutional Neural Networks (CNNs) is that they require large datasets to obtain adequate robustness; on small datasets, they are prone to overfitting. Many methods have been proposed to overcome this shortcoming with CNNs. In cases where additional samples cannot easily be collected, a common approach is to generate more data points from existing data using an augmentation technique. In image classification, many augmentation approaches utilize simple image manipulation algorithms. In this work, we build ensembles on the data level by adding images generated by combining fourteen augmentation approaches, three of which are proposed here for the first time. These novel methods are based on the Fourier Transform (FT), the Radon Transform (RT) and the Discrete Cosine Transform (DCT). Pretrained ResNet50 networks are finetuned on training sets that include images derived from each augmentation method. These networks and several fusions are evaluated and compared across eleven benchmarks. Results show that building ensembles on the data level by combining different data augmentation methods produce classifiers that not only compete competitively against the state-of-the-art but often surpass the best approaches reported in the literature.