Introduction
GenTO is an innovative project that introduces an interactive 3D virtual fitting display for fashion retailers. Utilizing advanced body measurement technology, MetaHuman models, and Unreal Engine, this system allows users to see themselves in virtual outfits in real-time by standing in front of a storefront display. The project aims to enhance customer engagement, drive in-store visits, and streamline the shopping experience by eliminating the need for physical try-ons.
Key Features:
Real-Time 3D Try-On: Captures user measurements and dynamically applies virtual outfits, powered by pose estimation and motion detection tools.
MetaHuman Integration: Creates realistic avatars for users, improving immersion and accuracy in the fitting process.
Face-Swapping Technology: Uses deepfake features to seamlessly integrate user facial features into their virtual avatars.
Efficient Backend: Unreal Engine provides real-time rendering and customization for a smooth, visually appealing user experience.
Development Progress:
Transitioned from a web-based solution to a storefront display for enhanced interactivity.
Drafted a business model canvas and conducted a 5-year NPV analysis to evaluate financial viability.
Implemented early-stage workflows, including survey creation and iterative design using Figma for UI/UX prototypes.
Challenges:
Rendering realistic 3D avatars remains computationally intensive, requiring high-performance GPUs.
Pose estimation and motion detection tools face limitations when tracking users on the move.
Ensuring face-swapping accuracy under varied lighting and orientations poses technical hurdles.
Future Goals: