The rose is one of the world’s most significant flowers, which is cultivated widely. But, it is susceptible to various types of diseases that can lead to a substantial reduction in yield. Hence, early detection of these diseases is essential to avoid risk and ensuring timely treatment. For this purpose, a deep learning-based approach has been proposed to detect rose leaf disease at an early stage. The proposed work uses two prominent variants of You Look Only Once (YOLO) model, i.e., YOLOv5 and YOLOv8, which is basically a Convolutional Neural Network (CNN)-based model. The experimental results show that, YOLOv8 model has achieved 71% precision, 65.1% recall, 67% F1 score, and 66.3% mean average precision at 0.5 (mAP(50). Whereas, YOLOv5 has achieved 70% precision, 60% recall, 65% F1 score, and 66.4% mAP(50). We have conducted extensive experiments on different dataset collected from both online resources and real field. The evaluation result shows that, the model has outperformed other existing models, and can be considered as the best choice for rose disease detection.