Improvement of Online Game Anti-Cheat System based on Deep Learning

With the popularity of video games and the increase of game players, cheating has become more common. Thus, game operators have designed and applied various anti-cheating systems. However, they each has some problems, the most general of which is the lag of penalty. In this article, AI models will be applied to improve anti-cheating systems. The paper first discusses the existing anti-cheating systems and their drawbacks, referring to some relevant literature. After that, two main scenarios are proposed, and one of them is actually modeled, trained and tested. This part is mainly done with python coding and the help of YOLOv5 program. The model still has room for improvement, but it does have substantiated the feasibility of above schemes.