Enhancing Sugarcane Leaf Disease Classification Using Vision Transformers over CNNs

Strategic Insights -

Researchers at the Department of Artificial Intelligence and Data Science have explored the use of Vision Transformers (ViT) for classifying sugarcane leaf diseases, demonstrating superior performance over traditional CNNs. The study utilized a dataset of 19,926 images across six classes to fine-tune both ViT and CNN models, achieving a test accuracy of 96.53% with the optimized ViT model. The results show the potential of ViTs in early disease detection for sustainable crop management.

Key Takeaways:

  • The study used a dataset of 19,926 images across six classes to fine-tune both ViT and CNN models.
  • The optimized ViT model achieved a test accuracy of 96.53%, outperforming the CNN models (ResNet50 and VGG16) with accuracies of 91.92% and 92.30%, respectively.
  • The ViT model demonstrated superior performance over CNNs in early disease detection for sustainable crop management.
  • The research concluded that future work will focus on expanding the dataset and optimizing model parameters for further improvements in disease classification accuracy.
  • The study was conducted by Saritha Miryala and Krupa Rasane from the Department of Artificial Intelligence and Data Science, S.G. Balekundri Institute of Technology.
  • The results have implications for the development of more efficient and accurate disease detection systems for sustainable crop management.

Statistics:

  • 96.53%: The test accuracy achieved by the optimized ViT model.
  • 19,926: The number of images in the dataset used to fine-tune both ViT and CNN models.
  • 6: The number of classes in the dataset.
  • 91.92% and 92.30%: The test accuracies achieved by the ResNet50 and VGG16 CNN models, respectively.

Sources:

  • Discover Artificial Intelligence, 2025,5(1):1-18
  • doi-org.sdpl.idm.oclc.org/10.1007/s44163-025-00340-7
  • NewsRx. Researchers at Department of Artificial Intelligence and Data Science Target Artificial Intelligence (Enhancing sugarcane leaf disease classification using vision transformers over CNNs). Robotics & Machine Learning. June 23, 2025; p 613.