Note: this is a past event Check out the current event!

Paris Buttfield-Addison

Paris Buttfield-Addison

Dr. Paris Buttfield-Addison is co-founder of Secret Lab a game development studio based in beautiful Hobart, Australia. Secret Lab builds games and game development tools, including the multi-award-winning ABC Play School iPad games, the BAFTA- and IGF-winning Night in the Woods, the Qantas airlines Joey Playbox games, and the Yarn Spinner narrative game framework. Previously, Paris was mobile product manager for Meebo, a ground-breaking 'Web 2.0' startup which was acquired by Google. Paris particularly enjoys game design, statistics, the blockchain, machine learning, and human-centered technology research and writes technical books on mobile and game development (more than 20 so far) for O’Reilly Media. He is currently writing 'Practical AI with Swift', 'Head First Swift', and the 'Unity Game Development Cookbook'. He holds a degree in medieval history and a PhD in computing.

Simulating your robot brain with a game engine

Level: General

The future of machine learning is big. Really big. Machine learning will be an important part of the software world for decades to come, in everything from robots, to health, to logistics, to the small and personal world. Machine Learning touches the real world all the time. This session explores how you can simulate enough of the real world with a video game engine to do machine learning. Join us.

Specifically, will explore:

• how game engines like Unity, Unreal, and Godot are on the forefront of ML by allowing us to simulate “just enough” of the real world to be useful

• how sophisticated machine learning frameworks like TensorFlow (and PyTorch) can be connected to a game engine to generate useful training data, train agents, and create robot brains which can then be transposed to the real world

• how machine learning requires so much data and infrastructure that it makes financial, business, and engineering sense to simulate things instead of spending money, time, and data scientists before you know if it worksWe’ll use the Unity game engine and the popular open source combination of the Unity ML-Agents Toolkit and TensorFlow as a case study to show how game engines can be used to perform agent-based deep reinforcement learning, behavioural cloning, and more, as well as generate simulated data for machine learning.