Journal DOI 10.68956/journal.human-performance-futures

Creative Commons source reading · Machine learning for performance

Enhancing multiclass plant disease classification using GAN-boosted vision transformer with XAI insights

Felicita S. A. M.; Kavitha B. R.

Originally published in Front Plant Sci · 2026-01-09

From the article

Selected source text. JournalOS removes markup and shortens excerpts where indicated. The full article provides the complete context.

Abstract

IntroductionAgriculture is one of the major backbones of the Indian economy, where rice is the most prominent staple crop across the country. However, rice production has been significantly affected due to the occurrence of various plant diseases. Deep learning and machine learning have emerged as powerful solutions for computer vision-based problems.MethodsThis work identifies some of the key diseases and addresses these prominent ones using a state-of-the-art deep learning model. It proposes a novel multiclass rice leaf disease recognition model named GRG-ViT, which integrates Vision Transformer (ViT), Generative Artificial Intelligence (GenAI), and Explainable Artificial Intelligence (XAI) techniques for better outcomes. The Vision Transformer-based framework is designed to capture long-range spatial dependencies in leaf images,…

Results and discussion

This section describes the evaluation results in detail with the classification performances of the proposed model. Three stages of comparison are performed to analyze the performance of the proposed work. First, it is compared with the performance of the ViT using an imbalanced dataset, then with the basic ViT architecture without any hyper-tuning for the balanced dataset, and last with the chosen pre-trained models of CNN.

Conclusion

Sustainable agricultural productivity is key to ensuring food security, and apart from weather-related threats, diseases cause major damage to productivity. This work presents an advanced deep learning model using Vision Transformer for more accurate disease classification. It addresses class imbalances through GenAI-based techniques, and the results show that the balanced dataset produces better outcomes compared to the imbalanced one. Model interpretability is another key feature of this work, bringing greater reliability to the results, which in turn supports possible real-time deployment of the model. The Grad-CAM and Attention Map visualizations provide evidence-based insights, which not only validate the results but also offer valuable information to agronomists. The key strength of the Vision Transformer is its self-attention…

Editorial reading guide

  • What data created and tested the model, and was evaluation independent?
  • What is the cost of a false positive, false negative or out-of-distribution decision?