Style AI: An Integrated AI-Driven Framework For Virtual Wardrobe Management And Fashion Try-On

Abstract

The rapid expansion of e-commerce has significantly transformed the fashion retail industry; however, customers still face challenges in visualizing garment fit and appearance before purchase. This limitation leads to high return rates and customer dissatisfaction. This paper presents Style AI, an integrated artificial intelligence-driven framework for virtual wardrobe management and fashion try-on. The system utilizes computer vision, pose detection, and machine learning algorithms to allow users to digitally try on clothing using their own images without requiring specialized hardware. Additionally, it provides wardrobe organization and personalized outfit recommendations based on user preferences, seasons, and occasions. The platform is developed using React-TypeScript for the frontend and FastAPI for the backend with cloud-based storage support. Experimental evaluation shows improved visualization accuracy, reduced uncertainty in purchasing decisions, and enhanced user satisfaction.