For the past several years the Unity has been steadily expanding its’ open-source plugin for training Machine Learning Agents (based on TensorFlow) in the Unity game engine. This talk will provide an overview of the capabilities of Unity’s ML Agent’s toolkit and the general AI principles underpinning it. With four years’ experience developing for XR in Unity, the speaker will also address tips and tricks for configuring and running the Machine Learning Agents plugin in Unity’s IDE. By the end of this talk, participants will have:
- Explored the capabilities of Unity’s ML Agent’s toolkit as an environment for training intelligent agents.
- Received an overview of the machine learning principles behind the various aspects of the toolkit.
- Explored potential use cases for training intelligent agents in the Unity game engine, with tips and tricks for working in Unity’s IDE.