A benchmark dataset of crop leaf images for image classification. Images are labeled across 38 crop disease classes, suitable for training agricultural vision models.
54,000 labeled images
38 disease classes
Train/validation/test splits
Use with standard image classification training pipelines.
