Skip to content

Repository files navigation

IBM – Mixed Reality Board and Toy Games with Granite 3 SLMs

Project Website

https://students.cs.ucl.ac.uk/2024/group9/

Project Overview

IBM, UCL, and MotionInput Games are collaborating to build a mixed reality board game engine for educational and family use. This project leverages various AI models from IBM Granite, all offline and embedded for use on devices. The game can host up to 6 players or teams and runs natively in Unity, supporting cross-platform play on mobiles, tablets, and laptops.

Key Features

  • Mixed Reality Board Game Engine: Supports multiple game types and custom board designs.
  • AI Integration: Utilizes GenAI for imagery, Yolo for object detection and tracking.
  • Cross-Platform Compatibility: Runs on mobiles, tablets, and laptops.
  • Offline AI Models: Ensures small, lightweight models for device integration.
  • Virtual Dice: Facilitates turn-based gameplay.
  • Object Recognition: Uses offline AI with Granite to recognize toys and small figures as avatars.
  • Scoreboards: Tracks and stores user winnings.

Game Types

  1. Q&A Game: Individual turns with questions to answer within a time limit.
    • Questions sourced from spreadsheets, generated by Granite, or curated topics.
    • AR board placement with thematic backgrounds.
  2. Buzz-Game: Fast-paced answering.
  3. Team Competition: Teams compete against each other.
  4. Object Interaction: AI-generated questions from held toys and book pages.

Development Approach

The software follows the V-Model software engineering approach, ensuring continuous testing and deployability. The project aims for public end-user deployment and hopes to enter the Times Educational Supplement.

License

This project is licensed under the MIT License.

About

Mixed Reality Board Game

Resources

Stars

1 star

Watchers

1 watching

Forks

Releases

Packages

Used by

Contributors

Languages