Home https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Technology https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ War Stories: How Forza Learned to Love Neural Networks to Train AI Drivers

War Stories: How Forza Learned to Love Neural Networks to Train AI Drivers



Produced by Justin Wolfson, edited by Shandor Garrison. Click here to print.

Once a startup, the Power franchise is now established within the pantheon of great racing games. The first installment was created in response to the Xbox Fantastic tourism, but with a healthy help from online multiplayer racing as well. Since then, it̵

7;s grown with Microsoft’s Xbox consoles with more realistic graphics and increasingly accurate track-focused physics Forza Motorsport series as well as evolving into open adventure (and even a trip to the Lego dimension) for Forza Horizon game.

If you’re one of the millions of people who’ve played a Forza racing game, you’re probably aware of the players’ AI opponents, called “Drivatars.” When the first Drivatars debuted in Forza Motorsport in 2005, they were a significant improvement over the NPCs we ran in other driving games, which often just followed the same pre-programmed route around the track. “It was a machine learning system on a hard drive that used a Bayesian Neural Network to record [racing] lines and characteristics know how someone drove a car, ”explains Dan Greenawalt, creative director of Power franchise in Turn 10 Studios in our latest War Stories video.

In fact, the technology originated from Microsoft Research’s outpost in Cambridge, England, where computer scientists began using neural networks to see if it was possible to get a computer to identify a Formula 1 driver by the way they drove through corners.

In recent years, Drivatars has evolved significantly – and not always in ways that players appreciate. In the first few games written for Xbox and Xbox 360 consoles, Drivatars lived entirely on the console’s local hard drive. As you let go of the various tracks in the game, the Drivatar system quietly used this data in the background to train the AI ​​you were racing against. But with the transition to Xbox One, Greenawalt’s team decided to use the cloud.

“Suddenly we went to Thunderhead, we had millions of laps within a week?” Says Greenawalt. “It was crazy.”

More data, more problems

Very quickly, the team at Turn 10 began to learn a lot about how different players drove their cars. Tracks are not uniform – they vary in height, width, corner profile and more – and it complicates the way the neural network tries to generalize a race line. Things get even more complicated when you drive towards other cars; Human players are much more aggressive towards AI drivers than other people Forza Motorsport 6, the games began to include a way to shift some of AI’s aggression out as a way to combat single-player games that became demolition derbies as AI tried to destroy humans at every opportunity.

In this video – recorded externally due to COVID-19 – Greenawalt goes into some detail in remembering how his team has handled these challenges over the years and even discusses the dreaded problem of “rubber bands” where AI cars brake a little to allow a human player to catch up after a spin. Even if you are not a fan of racing games, we think you will find this one quite captivating.

List image of Ars Technica


Source link