Next-Generation Augmented Reality: A Deep Learning–Driven Framework for Intelligent Visual Perception and Interaction

Abstract

The rise of e-commerce has transformed the retail sector, providing customers worldwide with convenience. Nevertheless, one of the largest problems with online shopping is that customers cannot physically touch and feel products before purchasing them, particularly in the fashion retail sector. Most customers are reluctant to purchase accessories such as glasses, jewelry, hats, and shoes because they are not sure how they look when worn. This creates increased return rates, decreased customer satisfaction, and lost revenue for consumers. The Try Before Buy system is an on-the-go application that fuses Augmented Reality (AR) with AI-based recommendations to enable users to try fashion accessories virtually in real time. The system supports real-time face tracking, precise 3D object superimposition, and an AI-based recommendation engine to recommend accessories according to user choice. The application also enables users to post the outcome of their try-ons on social media, fostering engagement and interaction. This article presents the design, methodology, implementation, and testing of the system. The research also delves into existing AR-based try-on solutions, establishes their limitations, and shows how the Try Before Buy system enhances the online shopping experience. The proposed system massively lowers the uncertainty of online shopping, increases customer confidence, and reduces product returns.