Sign in Subscribe
Food supply

Novel Framework for Corn Disease Detection Using Few-Shot Learning Achieves High Accuracy

Strategic Insights

— 1 min read

Research from the Indian Institute of Information Technology Kottayam has led to the development of a novel framework, GRCornShot, for detecting corn diseases using few-shot learning with prototypical networks. This research aims to address the challenges of deep learning methods, which require large quantities of data to train models for diagnosis and further classification. The proposed GRCornShot model effectively classifies healthy and corn diseases using a 4-way 2-shot, 3-shot, 4-shot, and 5-shot learning strategy, achieving impressive accuracy of 96.19%, 96.54%, 96.90%, and 97% respectively. The incorporation of the Gabor filter into the backbone network ResNet-50 enhances the classification performance by extracting texture features.

Key Takeaways:

  • The proposed GRCornShot model uses few-shot learning with prototypical networks to address the challenges of deep learning methods in detecting corn diseases.
  • The model achieves impressive accuracy of 96.19%, 96.54%, 96.90%, and 97% using a 4-way 2-shot, 3-shot, 4-shot, and 5-shot learning strategy.
  • The use of Gabor filter in the backbone network ResNet-50 enhances the classification performance by extracting texture features.
  • The proposed framework has the potential to provide a robust solution for detecting corn diseases precisely with minimal data requirements.
  • The research has been published in the journal Scientific Reports, and the authors include Ruchi Rani, Jayakrushna Sahoo, Sivaiah Bellamkonda, and Sumit Kumar from the Indian Institute of Information Technology Kottayam.

Statistics:

  • The GRCornShot model achieves an accuracy of 96.19%, 96.54%, 96.90%, and 97% using a 4-way 2-shot, 3-shot, 4-shot, and 5-shot learning strategy.
  • The model uses a backbone network of ResNet-50, which is enhanced by the incorporation of Gabor filter.
  • The research aims to provide a robust solution for detecting corn diseases precisely with minimal data requirements, addressing the challenges of deep learning methods.
  • The study is published in the journal Scientific Reports, with the citation: A novel framework GRCornShot for corn disease detection using few shot learning with prototypical network. Scientific Reports, 2025;15(1):26461.

Sources:

  • A novel framework GRCornShot for corn disease detection using few shot learning with prototypical network. Scientific Reports, 2025;15(1):26461. (https://www.nature.com/articles/s41598-025-33523-4)
  • NewsRx. Indian Institute of Information Technology Kottayam Reports Findings in Engineering (A novel framework GRCornShot for corn disease detection using few shot learning with prototypical network). Journal of Engineering. August 4, 2025; p 1406.

Read more