A Domain-Adapted Lightweight Ensemble for Resource-Efficient Few-Shot Plant Disease Classification

2025-12-15

AI Summary

To address plant leaf disease recognition under resource-constrained agricultural settings—characterized by few-shot learning and limited computational capacity—this paper proposes a lightweight few-shot learning framework. Our method innovatively integrates domain-adaptive MobileNetV2/V3 feature fusion with an attention-enhanced Bi-LSTM classifier, significantly improving cross-domain generalization and discriminative robustness. On the PlantVillage dataset under a 15-shot setting, our model achieves 99.72% accuracy, surpassing the state-of-the-art by 3.72 percentage points, while maintaining a compact footprint of only 40 MB and 1.12 GFLOPs. Furthermore, on the real-world field dataset Dhan Shomadhan (15-shot), it attains 69.28% accuracy—substantially outperforming existing approaches. This work establishes an efficient, deployable diagnostic paradigm for sustainable smart agriculture at the edge.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Adversarial Attacks & Robustness
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Accurate and timely identification of plant leaf diseases is essential for resilient and sustainable agriculture, yet most deep learning approaches rely on large annotated datasets and computationally intensive models that are unsuitable for data-scarce and resource-constrained environments. To address these challenges we present a few-shot learning approach within a lightweight yet efficient framework that combines domain-adapted MobileNetV2 and MobileNetV3 models as feature extractors, along with a feature fusion technique to generate robust feature representation. For the classification task, the fused features are passed through a Bi-LSTM classifier enhanced with attention mechanisms to capture sequential dependencies and focus on the most relevant features, thereby achieving optimal classification performance even in complex, real-world environments with noisy or cluttered backgrounds. The proposed framework was evaluated across multiple experimental setups, including both laboratory-controlled and field-captured datasets. On tomato leaf diseases from the PlantVillage dataset, it consistently improved performance across 1 to 15 shot scenarios, reaching 98.23+-0.33% at 15 shot, closely approaching the 99.98% SOTA benchmark achieved by a Transductive LSTM with attention, while remaining lightweight and mobile-friendly. Under real-world conditions using field images from the Dhan Shomadhan dataset, it maintained robust performance, reaching 69.28+-1.49% at 15-shot and demonstrating strong resilience to complex backgrounds. Notably, it also outperformed the previous SOTA accuracy of 96.0% on six diseases from PlantVillage, achieving 99.72% with only 15-shot learning. With a compact model size of approximately 40 MB and inference complexity of approximately 1.12 GFLOPs, this work establishes a scalable, mobile-ready foundation for precise plant disease diagnostics in data-scarce regions.

Problem

• Develops a lightweight few-shot learning method for plant disease classification
• Addresses data scarcity and computational constraints in agricultural settings
• Enhances classification accuracy in real-world environments with complex backgrounds

Innovation

• Lightweight MobileNet models as domain-adapted feature extractors
• Feature fusion technique for robust representation in few-shot learning
• Bi-LSTM classifier with attention for sequential dependencies and focus