Facial Feature Recognition System Development for Enhancing Customer Experience in Cosmetics

Abstract

Face recognition has gained increasing attention during recent years thanks to technology development, however, the majority of its application is limited to access and security, while a much wider potential is yet to be exploited. Based on the insight provided by the literature review the present work research objective has been defined as the development of a smartphone application applying neural networks to identify and classify two eye features: color and shape. The main methodology steps include eye shape and color classes definition and selection, dataset collection and preprocessing neural network development, interface flow definition and smartphone application deployment. The result is an integrated and interactive system that is able to make relevant and customized make-up suggestionS to the user, achieving satisfying performances in terms of user-friendliness and accuracy (73.89% and 74.58% for color and shape classification, respectively). The present study proposes three main findings: definition of eye shapes and color classification, development of a neural network system for eye feature classification with high accuracy performances, deployment of a user-friendly smartphone app for personalizing customer experience in cosmetics. Therefore, the main contribution of the present study is expanding potential face recognition applications as well as providing a successful example of customer value creation through creatively applying Face recognition and Neural Networks.