Analysis of Deep Demographics for Advertisement Recommendation

Abstract

Advertisements are a popular marketing strategy that shapes consumer perception and brand image. Consumers engage in outdoor advertising messages and traditional media advertisements. Understanding consumer behavior and interest in advertisements is crucial for developing effective marketing strategies. One study used computer vision techniques to analyze customer demographics, clothing preferences, and facial attention cues to extract comprehensive features from individuals and assess their attention toward advertisement displays. The methodology uses object detection models, such as YOLO, to track individuals in a scene, followed by a fashion detection model to identify clothing styles. The MiVOLO model predicts age and gender and creates a dataset for demographic analysis. ReginaFace is used for face detection and head pose estimation to gauge viewer engagement. This system helps retailers and advertisers tailor marketing strategies based on real-time customer data, providing insights into consumer preferences and interests. This enhanced customer engagement and sales.