A. Prada Gonzalez and M. Brendel, "Leveraging Key-Point Detection to Prevent Shoe Returns in Online Shopping", Proc. of 3DBODY.TECH 2023 - 14th Int. Conf. and Exh. on 3D Body Scanning and Processing Technologies, Lugano, Switzerland, 17-18 Oct. 2023, #39, https://doi.org/10.15221/23.39.

Title:

Leveraging Key-Point Detection to Prevent Shoe Returns in Online Shopping

Authors:

Andres PRADA GONZALEZ, Matthias BRENDEL

Footprint Technologies GmbH, Berlin, Germany

Abstract:

More than half of all shoes ordered online are returned.1 Different shoe size systems, a lack of standards and deviating models lead to a high level of uncertainty on the part of consumers. As a result, retailers lose up to 20% of their revenues and the environment is unnecessarily polluted by packaging and shipping. Furthermore, users spend an average of 32 minutes of their time returning a shoe. We have developed a solution that allows consumers to measure their feet at home with their smartphone by processing two images of their feet on a DIN-A4 sheet. To accurately recommend a size, two sets of data are required: the feet measurements and the shoe dimensions. This is why, on the one hand, a foot measuring algorithm was developed using two computer vision techniques: semantic segmentation and key point detection. This paper compares both methodologies and evaluates which one performs better on a distance similarity metric. On the other hand, a database was built up where the Last data, i.e., the inner shoe dimensions of the shoes per manufacturer, model and size, are stored. When measuring the foot in 2D, there are factors such as the point of view of the images or the image resolution that can influence the perfect foot measurement. Hence, a guided capturing system was developed to minimize them. Finally, to match them to the perfect fitting size, we performed in-person tests with the manufacturer's shoes over the last year to optimize the way of recommending them.

Keywords:

foot measurement, shoe size recommendation

Details/PDF/VIDEO:

Full paper: PDF
Presentation: VIDEO
Proceedings: 3DBODY.TECH 2023, 17-18 Oct. 2023, Lugano, Switzerland
Paper id#: 39
DOI: 10.15221/23.39

Copyright notice

© Hometrica Consulting - Dr. Nicola D'Apuzzo, Switzerland, hometrica.ch.
Reproduction of the proceedings or any parts thereof (excluding short quotations for the use in the preparation of reviews and technical and scientific papers) may be made only after obtaining the specific approval of the publisher. The papers appearing in the proceedings reflect the author's opinions. Their inclusion in these publications does not necessary constitute endorsement by the editor or by the publisher. Authors retain all rights to individual papers.

Note: click the + on the top left of the page to open/close the menu.