Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers

Common bean production in Tanzania is threatened by diseases such as bean rust and bean anthracnose, with early detection critical for effective management. This study presents a Vision Transformer (ViT)-based deep learning model enhanced with adversarial training to improve disease detection robust...

Full description

Saved in:
Bibliographic Details
Main Authors: Upendo Mwaibale, Neema Mduma, Hudson Laizer, Bonny Mgawe
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-08-01
Series:Frontiers in Artificial Intelligence
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/frai.2025.1643582/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849236557414793216
author Upendo Mwaibale
Neema Mduma
Hudson Laizer
Bonny Mgawe
author_facet Upendo Mwaibale
Neema Mduma
Hudson Laizer
Bonny Mgawe
author_sort Upendo Mwaibale
collection DOAJ
description Common bean production in Tanzania is threatened by diseases such as bean rust and bean anthracnose, with early detection critical for effective management. This study presents a Vision Transformer (ViT)-based deep learning model enhanced with adversarial training to improve disease detection robustness under real-world farm conditions. A dataset of 100,000 annotated images augmented with geometric, color, and FGSM-based perturbations, simulating field variability. FGSM was selected for its computational efficiency in low-resource settings. The model, fine-tuned using transfer learning and validated through cross-validation, achieved an accuracy of 99.4%. Results highlight the effectiveness of integrating adversarial robustness to enhance model reliability for mobile-based plant disease detection in resource-constrained environments.
format Article
id doaj-art-b9f2ff06973e4e5cb4032a4187b01582
institution Kabale University
issn 2624-8212
language English
publishDate 2025-08-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Artificial Intelligence
spelling doaj-art-b9f2ff06973e4e5cb4032a4187b015822025-08-20T04:02:13ZengFrontiers Media S.A.Frontiers in Artificial Intelligence2624-82122025-08-01810.3389/frai.2025.16435821643582Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision TransformersUpendo Mwaibale0Neema Mduma1Hudson Laizer2Bonny Mgawe3Computational and Communication Science and Engineering (CoCSE), The Nelson Mandela African Institution of Science and Technology (NM-AIST), Arusha, TanzaniaComputational and Communication Science and Engineering (CoCSE), The Nelson Mandela African Institution of Science and Technology (NM-AIST), Arusha, TanzaniaLife Sciences and Bio-engineering (LiSBE), The Nelson Mandela African Institution of Science and Technology (NM-AIST), Arusha, TanzaniaComputational and Communication Science and Engineering (CoCSE), The Nelson Mandela African Institution of Science and Technology (NM-AIST), Arusha, TanzaniaCommon bean production in Tanzania is threatened by diseases such as bean rust and bean anthracnose, with early detection critical for effective management. This study presents a Vision Transformer (ViT)-based deep learning model enhanced with adversarial training to improve disease detection robustness under real-world farm conditions. A dataset of 100,000 annotated images augmented with geometric, color, and FGSM-based perturbations, simulating field variability. FGSM was selected for its computational efficiency in low-resource settings. The model, fine-tuned using transfer learning and validated through cross-validation, achieved an accuracy of 99.4%. Results highlight the effectiveness of integrating adversarial robustness to enhance model reliability for mobile-based plant disease detection in resource-constrained environments.https://www.frontiersin.org/articles/10.3389/frai.2025.1643582/fullbean rustbean anthracnosedeep learningVision Transformers (ViT)adversarial attacksFast Gradient Sign Method
spellingShingle Upendo Mwaibale
Neema Mduma
Hudson Laizer
Bonny Mgawe
Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
Frontiers in Artificial Intelligence
bean rust
bean anthracnose
deep learning
Vision Transformers (ViT)
adversarial attacks
Fast Gradient Sign Method
title Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
title_full Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
title_fullStr Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
title_full_unstemmed Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
title_short Enhancing detection of common bean diseases using Fast Gradient Sign Method–trained Vision Transformers
title_sort enhancing detection of common bean diseases using fast gradient sign method trained vision transformers
topic bean rust
bean anthracnose
deep learning
Vision Transformers (ViT)
adversarial attacks
Fast Gradient Sign Method
url https://www.frontiersin.org/articles/10.3389/frai.2025.1643582/full
work_keys_str_mv AT upendomwaibale enhancingdetectionofcommonbeandiseasesusingfastgradientsignmethodtrainedvisiontransformers
AT neemamduma enhancingdetectionofcommonbeandiseasesusingfastgradientsignmethodtrainedvisiontransformers
AT hudsonlaizer enhancingdetectionofcommonbeandiseasesusingfastgradientsignmethodtrainedvisiontransformers
AT bonnymgawe enhancingdetectionofcommonbeandiseasesusingfastgradientsignmethodtrainedvisiontransformers