Experimenting With Artificial Intelligence: Programming Pathfinding Algorithms in C++ With Unreal Engine 5

Description

Let’s make something to take back home as a souvenir! Join us in a beginner-friendly lab session on pathfinding algorithms for AI gameplay mechanics. Together, we will visually compare pathfinding algorithms and see them in action as an open-world game.

Abstract

This hands-on lab session focuses on sharing the art and value of decision-making systems and pathfinding algorithms for machine learning applications. Understanding theoretical frameworks and heuristic strategies is necessary to build a foundational understanding of AI for research, games, and beyond. Through game development, we experience a snippet of AI by visualizing pathfinding algorithms, an entry point to exploring any AI domain.

Together, we visualize the applications and influence of real-time pathfinding through programmatically modeling with Metahumans in Unreal Engine 5 (UE5). This comes together as an open-world game compatible with their preferred platform including WebGL, iOS/Android, and VR headsets. Personal hardware is not required for participation but is recommended for playtesting.

Links:

Name	Description	Link
SIGGRAPH Labs Program	Lab session details and information	SIGGRAPH 2024
Labs Paper	Paper publication link	ACM DL
Course Website	Course website and information	Webiste
Lecture Slides	Tuesday's presentatioin	Keynote Presentation

Requirements

• Unreal Engine:
  • Unreal Engine 5.4.3
  • Unreal Neural Network Engine (NNE)
  • MetaHuman Plugin
  • MetaHuman Creator
  • Environment Query System
• Other:
  • Git Large File Storage
  • Kenny Assets: Survival Kit

Set Up

Fork or clone the entire repo

$ git clone https://github.com/debbieyuen/tag-youre-it

Credits and References