Under the hood of this powerful new suite of tools (download links can be found at the bottom of the page) resides the Intel® Distribution of OpenVINO™ toolkit, which simplifies the development process by helping target the right accelerators on the CPU, GPU, or both simultaneously. This enables game developers to harness the power of AI and machine learning with ease. It also leverages DP4a instructions in our latest integrated GPUs and is ready to take advantage of new XMX AI acceleration in our upcoming Intel® Arc™ graphics product family, too.
In addition, the AI GameDev Toolkit now enables you to apply generative adversarial networks (GANs) style transfer capabilities to stylize entire scenes or to focus on specific objects and characters within a scene, giving additional artistic flexibility.
One of the coolest new features the AI GameDev Toolkit brings to developers is the ability to unlock the potency of in-game object detection. Imagine enhancing your games to provide AI-assisted target identification, or providing mods that intelligently augment the game experience based on player preferences and learnings.
Also of note, Intel has partnered with game development asset-creator Procedural Worlds* to combine the AI inferencing power of the latest Intel® Graphics technology with their well-established and developer-friendly Gaia interface for Unity*. Gaia ML will integrate seamlessly with the AI GameDev Toolkit. Extending this AI power into Procedural Worlds’ next-generation Meta Worlds platform promises to elevate game-world-creation capabilities to a previously unimaginable level. You can now create richly textured worlds that feature complex and photo-realistic terrains, lush vegetation, and lifelike environmental systems—in mere minutes.
As you prepare to dive into the newly released AI GameDev Toolkit, check out three tutorials from Intel® Software Innovator Christian Mills that will help educate devs on the key components of the software.
AI Scene Style Transfer Plug-In with OpenVINO and Unity*
This three-part tutorial explores AI inferencing technology, and its usage in game development (and other creative outlets). In part one, Christian extended his previous series on in-game style transfer to encompass usage of the Intel Distribution of OpenVINO toolkit. Links to part two and part three are embedded in the tutorial, so you can study this computationally demanding style transfer model to appreciate the potential performance benefits of using the OpenVINO inference engine.
AI Style Transfer Plug-In for Unity, Enabling Targeted GameObject Stylization
This tutorial focuses on style transfer technology. Christian has extended his training in the machine learning and style transfer world to cover a method to stylize specific GameObjects in Unity. You’ll find step-by-step instructions, sample code, graphics, and examples.
OpenVINO Object Detection Inferencing Plug-In for Unity
In this tutorial Christian demonstrates how to leverage the Intel Distribution of OpenVINO toolkit in order to implement AI object-detection technology in games. Game developers can use these foundational tutorials to expand their knowledge of AI and machine learning, and learn to build even more amazing experiences.
These AI-enabled capabilities are just the beginning, as the Innovator Community explores new ML gaming features to further power the toolkit later this year. Intel plans on supporting Unreal* Engine in 2022, and adding more AI-enabled capabilities to the toolkit throughout the year.
or to take full advantage of what the AI GameDev Toolkit can enable