Home > Technology peripherals > AI > Gradually master the pinball maze skills: Novice turns into expert, CyberRunner AI speeds up the level

Gradually master the pinball maze skills: Novice turns into expert, CyberRunner AI speeds up the level

PHPz
Release: 2024-01-12 22:57:06
forward
543 people have browsed it

IT House reported on December 20 that researchers at ETH Zurich recently produced the CyberRunner robot, revealing the extraordinary learning capabilities of artificial intelligence (AI).

6小时从新手到专家,CyberRunner AI解锁通关弹珠迷宫技能

The research team used a large marble maze for testing. By avoiding falling into the holes, they followed the set maze path to reach the destination

The CyberRunner AI robot is equipped with two electric motors that drive the knobs, a camera to view the circuit board, and an AI brain running a "deep reinforcement learning system" capable of sending up to 55 control instructions per second. After rewriting: The CyberRunner AI robot is equipped with two motors with drive knobs, a camera for monitoring the circuit board, and an AI brain running a "deep reinforcement learning system" that can send up to 55 messages per second Control instruction

The movement of the ball can be controlled indirectly through two knobs, which can change the direction of the chessboard. Although this game is relatively simple, it requires fine motor skills and spatial reasoning. According to experience, humans need a lot of practice to master this game

6小时从新手到专家,CyberRunner AI解锁通关弹珠迷宫技能

After just 6.06 hours of training, CyberRunner quickly grew from a beginner to a smooth, experienced professional, even able to find ways to cheat like a human.

CyberRunner speeds up game play by more than 6% after beating the fastest record time of an "extremely skilled human player." The following is the video link attached to IT House:

The above is the detailed content of Gradually master the pinball maze skills: Novice turns into expert, CyberRunner AI speeds up the level. For more information, please follow other related articles on the PHP Chinese website!

source:sohu.com
Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn
Popular Tutorials
More>
Latest Downloads
More>
Web Effects
Website Source Code
Website Materials
Front End Template