SR-VTRON: A Web Application to Enhance the Online Retail Shopping Experience Through Augmented Reality and LightFM

Herrera R.; Tomayconza S.; Wong L. · 2026 9th International Conference on Artificial Intelligence and Big Data, ICAIBD 2026 (2026) · DOI: 10.1109/icaibd69640.2026.11637351

Resumen

The growth of online apparel retail has increased uncertainty regarding size and fit, leading to frequent returns and a decline in customer confidence. To address this issue, we propose SR-VTRON, a web-based virtual fitting solution that integrates augmented reality and machine learning techniques, enabling users to preview how a garment would look on their own image before making a purchase. The approach was developed in three stages: (i) selection of augmented reality libraries, (ii) selection of machine learning components based on LightFM, and (iii) construction of the web application. Validation was carried out with 40 users, examining the correct execution of the full interaction flow, perceived usability, and the reduction of uncertainty related to size and fit. The results showed a reduction in the return rate from 29.3% to 11.1%, along with a 93% user satisfaction rate, indicating that SR-VTRON performed the entire process without errors and produced visually consistent simulations-suggesting its potential to reduce product returns and strengthen consumer confidence in online apparel purchases.

Instituciones: Universidad Peruana de Ciencias Aplicadas