《西部闹市》模块包现已免费放出

MindMaker AI Plugin

Aaron Krumins - 11月 6, 2020

AI Plugin that enables Machine Learning within Unreal Engine

  • 支持的平台
  • 支持的引擎版本
    4.26
  • 下载类型
    引擎插件
    此产品包含一款代码插件,含有预编译的二进制文件以及与虚幻引擎集成的所有源代码,能够安装到您选择的引擎版本中,并根据每个项目的需求启动。

Video: https://youtu.be/ERm_pZhAPIA


The MindMaker AI Plugin is an open-source plugin that enables games and simulations within UE4 to function as environments for training autonomous machine learning agents. The plugin facilitates a network connection and subsequent launch of a compatible MindMaker Learning Engine, for instance the Deep Reinforcement Learning(DRL) Engine. Agents can currently be trained using deep reinforcement learning, a machine learning approach that combines neural networks with a learning model to sculpt agent behavior. With MindMaker, developers and researchers can easily train machine learning agents for 2D, 3D and VR projects.


Possible use cases include robotic simulation, autonomous driving, generative architecture, procedural graphics and much more. MindMaker AI Plugin provides a central platform from which advances in machine learning can reach many of these fields. For game developers, the use cases for self-optimizing agents include controlling NPC behavior (in a variety of settings such as multi-agent and adversarial), prototyping game design decisions, and automated testing of game builds.


Algorithms presently supported include : Actor Critic ( A2C ), Sample Efficient Actor-Critic with Experience Replay (ACER), Actor Critic using Kronecker-Factored Trust Region ( ACKTR ), Deep Q Network ( DQN ), Proximal Policy Optimization ( PPO ), Soft Actor Critic ( SAC ), Twin Delayed DDPG ( TD3 )


技术细节

Features:

  • Real-time bi-directional communication with a compatible MindMaker Learning Engine
  • Auto-launch of the MindMaker Learning Engine
  • Enables Cutting Edge Machine Learning Within UE4
  • Direct Access to Neural Networks via Blueprints
  • Two Example Projects Included
  • Modular Design for Easy Integration
  • Socket.IO v1.4+ support


Code Modules:

  • MindMakerBPLibrary - Developer, CoreUtility - Runtime, SIOJson - Runtime, SocketIOClient - Runtime, SIOJEditorPlugin – Developer,

Number of Blueprints: 0

Number of C++ Classes: 18

Network Replicated: No

Supported Development Platforms: Win64

Supported Target Build Platforms: Win64

Documentation: http://www.autonomousduck.com/plugin.html

Tutorial: Creating a Custom Deep Reinforcement Learning Agent in UE4

Example Project: https://drive.google.com/file/d/1CNMmyOUhQJCFDJjUI-I15nd3wwXm2wIY/view?usp=sharing

Important/Additional Notes: Download Learning Engine and Starter Content