Retail Store Analytics Using Facial Recognition

Abstract

—In this paper, we present a system for real-time age, gender, and emotion detection using webcam and machine learning techniques. The system is designed to capture real-time video footage of customers in a retail store and extract demographic and emotional information to perform retail analytics. We used the UTKFace dataset to train our age and gender model and the FER dataset to train our emotion model. The trained models were integrated with OpenCV and TensorFlow to detect faces, predict age, gender, and emotion in real-time. The system stores the collected data in a MySQL database, which is then used to perform various analyses to gain insights into customer behavior. We provided analysis using Flask and built a web interface for getting the insights on our data. The results show that our system is effective in capturing and analyzing customer information in real-time and can provide valuable insights for retailers to make informed decisions. Our system can be extended to include other features such as customer segmentation, heatmaps, and customer journey analysis. Overall, our system provides a powerful tool for retailers to understand customer behavior and improve the shopping experience.