Touchless Advertising Mobile Application - TaMa

Abstract

Hand gesture recognition has been a promising area of research in the past decades, especially with breakthroughs in the field of computer vision, but the COVID-19 pandemic has brought much attention to this field. The emphasis on a higher sanitary standard has pushed for more touchless interactions to help mitigate viral contagion. This technology could prove to be highly beneficial for interactive devices located in public spaces, such as self-serving information kiosks and self-service check-outs. This paper proposes a touchless advertising mobile application as a proof of concept, to display and interact with advertisements, as a proof of concept. The application uses an implementation of the TensorFlow library to detect, extract, and classify hand gestures. This system is tested and verified to show its robustness. The results obtained show favourable performance and accuracy. The application is designed to offer a smoother learning curve, making it easy to use.