Deep Learning and Optimization Approaches in IoT based soil nutrition and plant disease detection system for smart agriculture using Multi-Layer Stacked Residual Coordinate Boosted Sooty Tern Attention Network: A Review

Abstract

The integration of the Internet of Things (IoT) and Artificial Intelligence (AI) has significantly transformed modern agriculture by enabling intelligent systems for soil nutrition management and plant disease detection. Traditional farming practices often face challenges such as delayed disease identification, inefficient nutrient management, and limited real-time monitoring. To overcome these limitations, recent studies have explored advanced deep learning and optimization techniques for precision agriculture. This review examines deep learning approaches for IoT-based agricultural monitoring, focusing on the proposed Multi-Layer Stacked Residual Coordinate Boosted Sooty Tern Attention Network. The architecture combines residual learning, coordinate attention mechanisms, and attention-based feature enhancement to improve feature extraction and classification accuracy. Attention modules help identify critical regions in plant and soil data, while residual connections support deeper and more efficient network training. The study reviews recent research involving CNNs, CNN-LSTM hybrids, Vision Transformers, Graph Neural Networks, and optimization techniques such as genetic algorithms and swarm intelligence. These methods improve disease detection accuracy, resource utilization, and real-time monitoring. However, challenges including computational complexity, scalability, data dependency, and energy limitations remain significant for real-time IoT-enabled agricultural deployments.