Cross-Architectural Mixture-of-Experts with Adaptive Soft Routing for Plant Leaf Disease Classification

2026-06-22

AI Summary

Plant leaf disease classification remains challenging under complex backgrounds, varying illumination conditions, and class imbalance, as single models often fail to effectively capture both local and global features. This work proposes an adaptive soft-routing mixture-of-experts (MoE) framework, introducing cross-architecture MoE for the first time to plant disease identification by integrating EfficientNet-B0, DenseNet-121, and Swin-Tiny. The model employs dynamic gating to enable complementary multi-scale feature representation and adopts a two-stage training strategy to enhance stability and generalization. Evaluated on multiple datasets, the method achieves F1 scores of 92.62% on potatoes—surpassing the best single model by over 5%—and further demonstrates strong performance with 94.03% on durian and 97.04% on sesame, confirming its effectiveness and robust generalization capability.

Abstract

Plant leaf disease classification is crucial for crop protection and precision agriculture but remains challenging under complex backgrounds, illumination variations, and severe class imbalance. Moreover, single-architecture models often fail to effectively capture both local and global representations. To address these challenges, this study proposes an adaptive soft Mixture-of-Experts (MoE) framework with cross-architectural routing that integrates EfficientNet-B0, DenseNet-121, and Swin-Tiny to exploit complementary multi-scale, local, and global features. A soft gating mechanism dynamically assigns input-dependent expert weights, while a two-stage refinement training strategy improves optimization stability and generalization. Experiments on a highly imbalanced potato leaf disease dataset achieve 91.68% recall and 92.62% F1-score, surpassing the strongest individual expert by 5.91% and 5.03%, respectively. Additional evaluations on durian and sesame leaf disease datasets yield F1-scores of 94.03% and 97.04%, demonstrating robust cross-dataset generalization and the potential of the proposed framework for reliable real-world crop health monitoring

Problem

• plant leaf disease classification
• class imbalance
• complex backgrounds
• illumination variations
• local and global representations

Innovation

• Mixture-of-Experts
• Cross-Architectural Routing
• Adaptive Soft Gating
• Multi-scale Feature Fusion
• Imbalanced Classification