<![CDATA[Leaf diseases in chili plants, such as leaf spot and yellow virus, pose a significant threat to agricultural productivity in Indonesia, often leading to substantial economic losses for farmers. Traditional manual identification remains inefficient and highly dependent on individual expertise, which frequently results in inconsistent diagnosis. This research proposes an automated detection system utilizing the YOLOv8n deep learning architecture to provide a more reliable and real-time solution. A major hurdle in developing robust AI models for agriculture is the scarcity of balanced field dataset s and the presence of complex natural backgrounds. To address this, the study employs Leafgan , a generative augmentation technique capable of transforming healthy leaf images into realistic diseased samples while preserving the original field environment. By leveraging an attention mechanism, Leafgan maintains high-frequency textural details, allowing the YOLOv8n model to generalize better across diverse environmental conditions. Data management is streamlined through the Roboflow platform to ensure consistent integration of primary and synthetic dataset s. The primary goal of this integration is to enhance model stability, aiming for a minimum mean Average Precision (mAP) within actual plantation settings.]]>
Copyrights © 2026
