AI Enabled Virtual Try-On for E-Commerce

In today’s digital environment, online fashion sales face return rates of 30 to 40 percent because customers cannot try products or things before buying. This paper introduces StyleGenie, an AI-enabled virtual try-on system that allows users to see photorealistic product images on their own photos The software uses Google Gemini generative multimodal AI model architecture for generating realistic images in more than nine categories of products including eyewear, shirts, hoodies, t-shirts, watches, women’s tops, and traditional kurtas. A multi-model fallback strategy achieves an average try-on generation success rate of 92.9 percent. The platform offers AI-driven style analysis, which includes face shape detection, fit scoring, color harmony evaluation, and a stylist’s recommendation. Built on React 18 with TypeScript and a secure Node.js/Express backend proxy, the system prioritizes user privacy and does not store images permanently. Client-side image processing reduces file size by over 70 percent. A pilot evaluation with 35 participants results in a System Usability Scale (SUS) score of 84.2, which is considered ‘Excellent,’ along with a 76 percent self-reported decrease in purchase uncertainty.