Customer Activity Recognition in Retail (CARR) Dataset

The Customer Activity Recognition in Retail (CARR) dataset is a publicly available dataset designed to support research and development in customer behavior analysis using video surveillance. The dataset includes annotated videos and frames of customer activities captured in realistic retail environments using a single surveillance camera with a tilted angle of 45–90°.This dataset is intended for training and evaluating machine learning and deep learning models—particularly object detection, pose estimation, and activity classification—in retail contexts.