Dataset for: Semi-Supervised Deep Learning for Rice Leaf Disease and Pest Classification via IoT

This dataset contains the image data used in the study "Semi-Supervised Deep Learning for Rice Leaf Disease and Pest Classification: Leveraging 10,993 Unlabeled Field Images Acquired via IoT". The dataset is split into two parts: Expert-Labeled Dataset (744 images): Verified by three plant pathology and breeding experts from Universitas Jenderal Soedirman (UNSOED). It contains 10 classes (7 diseases, 2 pests, and 1 healthy class).Unlabeled Dataset (10,249 images): Raw field images collected directly from rice fields using an ESP32-CAM-based IoT device under varying real-world conditions (illumination, angles, growth stages).This dataset enables research in semi-supervised learning, pseudo-labeling, and edge-AI applications in smart agriculture.