Development of a Smart Digital Advertisement Board Based on Face Recognition System

Fahri Heltha (Universitas Syiah Kuala, Banda Aceh, Indonesia UCSI University, KL, Malaysia)
Sharandhass Radakrishnan (UCSI University, KL, Malaysia)
Haris Wahyudi (Department of Mechanical Engineering, Faculty of Engineering, Universitas Mercu Buana)
Aulia Rahman (Universitas Syiah Kuala, Banda Aceh, Indonesia)

International Journal of Advanced Technology in Mechanical, Mechatronics and Material (IJATEC)
Vol 4, No 1 (2023)

Abstract. We develop a smart digital advertisement board system which allows the system to display advertisements based on the majority of age and gender classifications of the consumers. The system captures the faces of the crowd and the face recognition techniques used to classify the majority gender and age of the crowd and then shows appropriate advertisement from the database to the advertisement board. A DNN model that is built, trained, and validated is used to recognize and predict the age and gender of the visible faces through image input or webcam using face photo dataset known as audience dataset. Several testing and analysis have been done onto the system in order to demonstrate the effectiveness and reliability of the system in displaying suitable advertisement for the public. The system can get gender accuracy of 77.82% and 86% for female and male respectively. And 68.78% accuracy for age recognition. The recognition speed is less than 1.3 second for up to 9 faces in an input image.