Kuan Ying Wu

RealityCast explores a lightweight workflow for bringing AI-generated characters into real-world AR experiences. The prototype combines generative AI, animation processing, and WebAR into a rapid production pipeline. A 2D chibi character is generated with AI, converted into a seamless looping animation, automatically separated from its background, and deployed as an interactive AR asset that runs directly in the browser. The workflow demonstrates that AI-generated animated characters can already function as practical, lightweight AR content without requiring traditional character modeling or rigging pipelines. By significantly reducing production time, RealityCast enables rapid prototyping and scalable deployment for interactive experiences. Potential applications include product visualization, location-based character experiences, tourism, promotional campaigns, and social media content, where virtual characters can be placed naturally into real-world environments through WebAR.