Learning Size and Fit from Fashion Images

Finding clothes that fit has increasingly become a hot topic in the e-commerce fashion industry. Mainly due to causing frustration on the customers side, and the large ecological and economical footprint on the companies side. Most approaches addressing this problem are based on statistical methods relying on historical data of articles purchased and returned to the store. Such approaches suffer from the cold start problem for the thousands of articles appearing on the shopping platforms every day, for which no prior purchase history is available. We propose to employ visual data to infer size and fit characteristics of fashion articles. We introduce SizeNet, a weakly-supervised teacher-student training framework that leverages the power of statistical models combined with the rich visual information from article images to learn visual cues for size and fit characteristics, capable of tackling the challenging cold start problem. We demonstrate the strong advantage of our approach through extensive experiments performed on thousands of textile garments, including dresses, trousers, knitwear, tops, etc. from hundreds of different brands.