Age and Gender Recognition for Masked Face Using YOLO-X and CNN in Smart Advertisement Systems

The conventional advertisement board often fails to attract its target customers effectively due to its limited ability to display content relevant to viewers. To address this, a Smart Personalized Advertisement (SAVER) board employing an age and gender recognition system is proposed. In the post-pandemic era, where many people wear face masks, developing effective smart advertising systems has become even more challenging. The research aims to evaluate and compare Convolutional Neural Network (CNN) architectures integrated with You Only Look Once-X (YOLO-X) for age and gender recognition in smart advertising applications that accommodate both masked and unmasked faces. The proposed framework first detects faces in an image using the YOLO-X model. The detected faces are then cropped based on bounding boxes and aligned to ensure consistent orientation. Subsequently, CNN classifies age groups and gender based on facial attributes. The detection results are used to determine which advertisements should be displayed. The research uniquely addresses the recognition of age and gender for both masked and unmasked faces and implements the solution in a realtime advertising system. The proposed system achieved 68% precision in delivering smart personalized advertisements, demonstrating its effectiveness in real-world public settings. In summary, this research contributes to the development of intelligent public display systems capable of delivering demographically aware content.