SERI-SENSE: A Comprehensive Machine Learning–Based Disease Identification and Recommendation System for Sericulture

Sericulture is a vital agro-based industry that significantly contributes to rural employment and the global silk economy. The productivity and sustainability of this sector are highly dependent on the health of silkworms and mulberry plants. Diseases affecting either silkworms or mulberry leaves can cause severe economic losses due to reduced cocoon yield, inferior silk quality, and increased mortality rates. Conventional disease identification methods rely primarily on manual visual inspection by farmers or experts, which is time-consuming, subjective, and often ineffective in detecting diseases at an early stage. This paper presents SERI-SENSE, a comprehensive machine learning–based disease identification and recommendation system designed specifically for sericulture applications. The proposed system integrates image processing techniques with a deep learning framework based on a pre-trained ResNet50 convolutional neural network to accurately classify diseases in silkworms and mulberry leaves. A transfer learning strategy is employed to improve performance and generalization with limited datasets. The system provides confidence-based disease predictions along with actionable preventive and corrective recommendations. Experimental evaluation demonstrates that SERI-SENSE achieves high classification accuracy and robustness, making it suitable for real-world deployment in precision sericulture.