Abstract:
FitMate is an innovative solution to the challenges of online clothing shopping, offering personalized fashion recommendations to enhance user experience. With the fashion industry increasingly shifting towards e-commerce platforms, customers face difficulties in finding clothes that fit well and form cohesive outfits, leading to high return rates and dissatisfaction. FitMate addresses this by leveraging algorithms to provide accurate clothing recommendations. The Body Category Detecting model employs KMeans clustering, an unsupervised method, to match user body measurements with clothing sizes, streamlining online shopping by eliminating the need for frequent size chart consultations. The Outfit Evaluator model within FitMate draws inspiration from Siamese Neural Networks, tailored to identify patterns when images of clothes are presented together. Unlike traditional Siamese Neural Networks, which measure similarity between two images, this model focuses on understanding the relationship and coherence between outfit items. By doing so, FitMate ensures that recommended outfits not only fit well individually but also complement each other aesthetically. FitMate aims to revolutionize online shopping by enhancing user satisfaction, reducing return rates, and improving the overall efficiency of fashion retail. Through its personalized approach and accurate recommendations, FitMate ensures to redefine the online shopping experience, making it more enjoyable and convenient for users.
