GG WORLDS:
A Blender plugin that enables users to instantly generate and customize 3D environments from text prompts, using generative AI to save time, cost, and boost creative freedom.
Physical Prototype:
Our prototype demonstrates an interactive workflow where users arrange physical maps and landmarks, then use a tablet AR app to instantly visualize and modify AI-generated 3D environments in real time. Bridging hands-on creativity with digital flexibility.
Target Audience:
People who want to showcase their digital products easily to make them understandable and grab attention.
- Game Developers
- Designers and 3D artists
- Researchers
Technical Details
User Journey
Colored objects representing landmarks are moved within the map area in front of a Raspberry Pi camera. Object recognition is performed using color-based detection. The coordinates of each detected landmark are then transmitted to the client application via an HTTP server, enabling accurate placement within the city model.
Dynamically generating 3d geometries by changing the base image.
The city model, complete with overlaid street networks, is visualized within the augmented reality application.
Future Developments
- Currently, the prototype uses QR codes to update elements. In the future, we aim to process live camera feeds, convert them to grayscale images, and use this data to generate terrain geometry in real time.
- We also plan to make buildings automatically adapt to the positions of landmarks.
