Image augmentation improves few-shot classification performance in plant disease recognition

TL;DR

This paper demonstrates that applying a sequence of data augmentation techniques significantly enhances the accuracy of plant disease classification models trained with very limited data, using transfer learning with ResNet.

Contribution

The study introduces an augmentation scheme that combines multiple techniques to improve few-shot plant disease classification performance.

Findings

Augmentation scheme increases accuracy by over 25% with only 10 seed images.

Combining multiple augmentation techniques yields better results than individual methods.

The approach is effective in scenarios with scarce labeled data.

Abstract

With the world population projected to near 10 billion by 2050, minimizing crop damage and guaranteeing food security has never been more important. Machine learning has been proposed as a solution to quickly and efficiently identify diseases in crops. Convolutional Neural Networks typically require large datasets of annotated data which are not available on demand. Collecting this data is a long and arduous process which involves manually picking, imaging, and annotating each individual leaf. I tackle the problem of plant image data scarcity by exploring the efficacy of various data augmentation techniques when used in conjunction with transfer learning. I evaluate the impact of various data augmentation techniques both individually and combined on the performance of a ResNet. I propose an augmentation scheme utilizing a sequence of different augmentations which consistently improves accuracy through many trials. Using only 10 total seed images, I demonstrate that my augmentation framework can increase model accuracy by upwards of 25\%.