Authors:

Sadique Nayeem

Abstract:

Rice is a staple food for billions, and its yield is persistently threatened by a spectrum of diseases. The timely and accurate diagnosis of these ailments is crucial for global food security. Recently, deep learning (DL) has emerged as a powerful tool to automate and enhance plant disease identification. This paper presents a systematic review of contemporary research applying DL models for rice disease diagnosis. Our analysis synthesizes findings from numerous studies, highlighting dominant architectural trends such as Convolutional Neural Networks (CNNs), transfer learning strategies, and the integration of data sources beyond standard RGB images. We critically evaluate the reported performance of these models, identifying common strengths and persistent challenges, including issues related to real-world variability, dataset limitations, and model generalizability. Finally, the review outlines promising future research directions, emphasizing the potential of multimodal data fusion, explainable AI (XAI) for building user trust, and the development of lightweight models suitable for deployment on mobile devices in resource-constrained agricultural settings.

Keywords:

Artificial Intelligence (AI), Machine Learning, AI-driven systems, Vector Machines (SVMs), Gray-Level Co-Occurrence Matrices (GLCMs), Convolutional Neural Networks (CNNs), and Artificial Neural Networks (ANNs).

Vol & Issue:

VOL.18, ISSUE No.1, March 2026