ABSTRACT Accurate identification and classification of rice leaf diseases are crucial for enhancing both the yield and quality of rice production. In this study, we evaluate the effectiveness of the EE‐YOLO11n model for classifying six types of rice leaf conditions: bacterial leaf blight, brown spot, leaf blast, narrow brown spot, leaf scald, and healthy leaves. Built upon the YOLO framework, our approach incorporates an innovative CSEE (Crop‐Specific Edge Enhancement) module and ADAC (Adversarial Data Augmentation for Crop diseases), to improve detection sensitivity to disease‐related visual features. We train and assess EE‐YOLO11n model on a publicly available rice leaf dataset, optimizing its training configuration and to balance high performance with computational efficiency. The configuration with a batch size of 32 achieves the best performance, yielding an accuracy of 0.998, which outperforms existing methods by approximately 4% in accuracy on the same dataset. Moreover, the proposed EE‐YOLO11n demonstrates superior results compared to state‐of‐the‐art approaches reported in the literature.
Zhang et al. (Wed,) studied this question.