Advanced Fashion Recommendation System for Different Body Types using Deep Learning Models

Abstract The fashion industry is rapidly expanding and playing a critical role in driving global economies. Due to this ever-growing industry, application of computer science is rising rapidly to solve different problems in this industry. Many e-commerce sites around the world allow their customers to purchase clothing items over the internet predominantly using recommender systems for shoppers based on the customer's purchase history, similar buying patterns of other shoppers, items in the wishlists and latest trends. These recommendation models lack personalization based on the user's body demographics. Since fashion is a way, one chooses to express themselves, it is important that each piece is carefully selected to suit the buyer. In this paper, an improved recommendation system is developed using a deep learning model for customers with different body shapes/types. It helps users to select clothing items based on their body shape. Proposed system is evaluated with respect to multiple deep learning models as well as traditional machine learning approaches. Xception model out performed by achieving 94% accuracy and a loss of 0.02%.