A Complexion based Outfit color recommender using Neural Networks
Reeta Koshy, Anisha Gharat, Tejashri Wagh, Siddesh Sonawane
Abstract
Shopping apparels and outfits either online or in stores is always a big task as it requires people to choose the most suitable outfits that resemble the latest fashion trends, have the perfect blend of outfit color combinations which matches their skin tone and looks as well. Thus confused customers need an advisor who can guide them through this outfits selection process. The aim of this paper is to propose a system which serves as a personalized fashion advisor. To serve this purpose, "Pocket Fashionista" which is a personalized Fashion Advisor is introduced that provides the best outfit and color combination recommendations to the users based on their complexion which will make them look great and feel confident about their appearance. The users can also visualize these recommended outfits on their own body to understand how the outfits and color combinations look on them thus enhancing their user experience. Weather based seasonal recommendations are also provided to the user. Hence, the system will help the E-commerce shopping websites, merchants and sales executives to master the real time demand of customers.