Plant Disease Classification with Vision Transformers and a CNN–Transformer Score Ensemble

DOI

10.2991/978-94-6239-799-6_7How to use a DOI?

Keywords

Plant Disease Classification; Vision Transformer; VGG16; Convolutional Neural Network; Ensemble Learning; Score Fusion; Precision Agriculture

Abstract

This paper examines Vision Transformer (ViT) models for classifying leaf diseases in cinnamon, rice, potato, and maize images. We compare a ViT initialized from ImageNet-21k weights with a custom ViT trained from random initialization. The custom model has approximately 399 million parameters, making it a high-capacity experimental architecture rather than a model intended for resource-constrained deployment. We also evaluate an equal-weight score ensemble of the scratch-trained ViT and a pretrained VGG16 network on the two-class cinnamon task. In the detailed cinnamon evaluation output, the scratch-trained ViT obtained about 84% accuracy and a macro F1-score of 0.84. Averaging its class probabilities with those produced by VGG16 yielded 89.57% accuracy and a macro F1-score of 0.90. The ensemble substantially reduced RoughBark false negatives, although StripeCanker recall declined slightly. Its overall accuracy was essentially equal to the separately reported pretrained-ViT result for cinnamon. The evidence therefore supports a narrower conclusion: convolutional predictions improved the recorded scratch-ViT result, but the ensemble did not outperform the strongest transformer baseline.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Divyesh Shah
AU  - Arun Warrier
AU  - Akshay Vhatkar
AU  - Pooja Raundale
PY  - 2026
DA  - 2026/09/30
TI  - Plant Disease Classification with Vision Transformers and a CNN–Transformer Score Ensemble
BT  - Proceedings of the International Conference on Emerging Trends and Technologies in Applications of Computer Science (ICETTACS 2026)
PB  - Atlantis Press
SP  - 87
EP  - 103
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-799-6_7
DO  - 10.2991/978-94-6239-799-6_7
ID  - Shah2026
ER  -