High-Accuracy Maize Disease Detection Based on Attention Generative Adversarial Network and Few-Shot Learning.

Area of Science:

• Agricultural Science
• Computer Vision
• Machine Learning

Background:

• Maize disease detection is crucial for agricultural production.
• Few-shot learning presents challenges due to limited data.
• Accurate disease identification impacts crop yield and management.

Purpose of the Study:

• To propose a high-accuracy maize disease detection method using Attention-GAN and few-shot learning.
• To enhance model performance by enabling focus on significant image regions.
• To overcome data scarcity issues in practical agricultural applications.

Main Methods:

• Utilized an Attention Generative Adversarial Network (Attention-GAN) for disease detection.
• Incorporated an attention mechanism to improve model focus on salient image features.
• Employed Generative Adversarial Network (GAN) for data augmentation to create synthetic training samples.

Main Results:

• Achieved high performance metrics: 0.97 accuracy, 0.92 recall, and 0.95 mean average precision (mAP).
• Demonstrated superior performance compared to baseline models in maize disease detection.
• Validated the method's accuracy and stability in few-shot learning scenarios.

Conclusions:

• The Attention-GAN based method offers a robust solution for maize disease detection with limited data.
• This approach advances agricultural informatization and intelligent farming practices.
• Provides a valuable reference for future research in agricultural disease diagnostics.