PPE Adherence Classification Dataset

This dataset is designed for training a PPE adherence classification network, which predicts the adherence status of healthcare providers based on their visual appearance. The dataset contains 12 classes corresponding to four types of personal protective equipment (PPE) — gown, mask, eyewear, and glove — each labeled according to adherence conditions. All images were cropped and sampled from the R2PPE dataset, followed by post-processing to isolate individual providers and normalize image dimensions.