Using Unity's Machine Learning Agents on Arm for Game Character Training

Explore Unity's Machine Learning Agents (ML-Agents) framework in this comprehensive 1-hour 15-minute workshop. Learn how to leverage this open-source intelligent game character kit for training single- and multi-agent scenarios using reinforcement learning, imitation learning, and other deep learning algorithms. Discover the framework's applications in games and apps for both mobile and PC platforms, built on top of Unity Barracuda. Gain insights into the practical implementation of ML-Agents on Windows on Arm, as demonstrated through the training of a boss character for the Dr Arm game using adversarial self-play and reinforcement learning. Investigate the potential of Arm CPUs for machine learning in games and explore ongoing research into multi-agent scenarios, including workload scaling and agent interactions. Prepare for the workshop by installing Unity, creating a GitHub account, and having an Android-based mobile phone with a USB connector. Access workshop files and additional resources through the provided session page link.