Building open-world AI that scales from large numbers of simple zombies to more intelligent bosses is an interesting challenge! This interview with Maciej Kurowski from Techland, covers the Hierarchical Task Network behind the game's AI. In particular, how it's connected to the sensory system, the networking of plans, why in-place planning was used, and how traversal works in the world.
You can now access the AiGameDev.com repositories on GitHub. These are used to various prototypes, libraries and other source code relating to your membership and upcoming conferences. Here's how it works...
Keen to experiment with coordinated search? The source code for the coordinated search tutorial is now available, along with bonus access to The AI Sandbox. You'll find a simple tactical pathfinder as well as a corridor map implementation, written in Python with extensive comments.
In this introductory tutorial about computer vision, you'll learn about homographic transformations, why they are useful and how they work. After that, you'll see how they can be applied to detecting the position of a camera in space, or the relative position of a projector. Then you'll also hear about useful computer vision algorithms, and why performance is often critical for such augmented reality ...
Crowd and ambient life simulation are having a huge impact on the visual effects industry. Many tools and and techniques are required to animate convincing crowds in films (Drakula Untold), TV series (The Walking Dead) and adverts. This interview digs into the technology using examples, and how these ideas and tools can be applied within real-time simulations and games.
Both commonly used techniques for simulating movement in groups have their challenges. Steering does not look particularly human-like, and velocity-based approaches like ORCA don't scale very well to larger crowds and have their own challenges.
This recording digs into most important ideas and techniques in Game AI from GDC last month, including the AI Summit, as well as big trends take-aways from the conference. You'll hear about the AI, animation and procedural generation in games such as Sunset Overdrive, Watch_Dogs, Galak-Z, and more.
Using Python is one of the simplest ways to get started with deep learning and neural networks. This tutorial will show you how you can get up and running with PyLearn2 via a simple wrapper library, and the various issues and questions you need to be considering in practice to get things working efficiently and reliably.
Keen to experiment with tactical pathfinding? The source code for the flanking tutorial is now available, along with bonus access to The AI Sandbox. You'll find a grid-based tactical pathfinder as well as one based on a corridor map, written in Python with extensive comments.
Over the past few years, deep learning has revolutionized the technology sector by providing significantly better results on classification or prediction problems that were previously impractical for machine learning. What are the benefit for game developers? What's the catch? This presentation is a practical introduction to deep neural networks, using millions of public player profiles and thousands of games ...