Virtual Jewellery Try-On: Enhancing Consumer Engagement and Personalization

Abstract

This study presents a mobile-based augmented reality (AR) solution for virtual jewelry try-on, designed to enhance the e-commerce user experience by enabling customers to visualize and interact with jewelry in real time. The system employs Google ML Kit for real-time pose detection and tracking of body landmarks such as shoulders and fingers. By integrating Flutter-based modules with machine learning models, the application achieves accurate overlay of virtual jewelry on detected positions. Interactive gesture-based controls, including tap-to-place, scaling, and rotation, allow users to customize the jewelry’s fit and appearance. To address challenges related to camera instability, smoothing algorithms are implemented to minimize visual jitter, ensuring a seamless experience. Preliminary evaluations indicate that the proposed solution delivers precise and responsive virtual jewelry placement, effectively engaging users with its intuitive interface and interactive features. Although certain limitations, such as manual adjustments and device performance constraints, persist, the system demonstrates strong potential for advancing digital jewelry retail. This research uniquely combines real-time AR, pose detection, and gesture recognition in a mobile application, offering an accessible and realistic try-on experience. It addresses critical usability challenges in the jewelry industry, bridging the gap between traditional shopping methods and digital innovation. This study highlights new opportunities for enhancing customer engagement and driving e-commerce growth within the jewelry sector.