Self-supervised transformer-based pre-training method with General Plant Infection dataset

TL;DR

This paper introduces a large, diverse plant pest and disease dataset and a novel transformer-based pre-training method combining Contrastive Learning and Masked Image Modeling to improve detection accuracy in agriculture.

Contribution

It presents a comprehensive dataset and a new pre-training approach that enhances plant pest and disease recognition performance.

Findings

Achieved high detection accuracy on the dataset.

Demonstrated the effectiveness of the combined Contrastive Learning and MIM approach.

Provided publicly available code and dataset for research advancement.

Abstract

Pest and disease classification is a challenging issue in agriculture. The performance of deep learning models is intricately linked to training data diversity and quantity, posing issues for plant pest and disease datasets that remain underdeveloped. This study addresses these challenges by constructing a comprehensive dataset and proposing an advanced network architecture that combines Contrastive Learning and Masked Image Modeling (MIM). The dataset comprises diverse plant species and pest categories, making it one of the largest and most varied in the field. The proposed network architecture demonstrates effectiveness in addressing plant pest and disease recognition tasks, achieving notable detection accuracy. This approach offers a viable solution for rapid, efficient, and cost-effective plant pest and disease detection, thereby reducing agricultural production costs. Our code and dataset will be publicly available to advance research in plant pest and disease recognition the GitHub repository at https://github.com/WASSER2545/GPID-22