{"id":1930,"date":"2023-04-17T13:35:53","date_gmt":"2023-04-17T17:35:53","guid":{"rendered":"https:\/\/sites.miamioh.edu\/cse-senior-design-expo\/?p=1930"},"modified":"2023-04-24T15:33:06","modified_gmt":"2023-04-24T19:33:06","slug":"aifps-game-defense-2023-team1","status":"publish","type":"post","link":"https:\/\/sites.miamioh.edu\/cse-senior-design-expo\/2023\/04\/aifps-game-defense-2023-team1\/","title":{"rendered":"AI-FPS video game defense Team 1"},"content":{"rendered":"\n\n\n \n \n \n \n \n \n
CLIENT<\/b><\/font><\/td>\n Dr Xianglong Feng, Assistant Professor
Computer Science & Software Engineering<\/td>\n <\/tr>\n<\/thead>\n
\n VISION<\/b><\/font><\/td>\n Recently, many have leveraged the power of computer vision algorithms and models to create cheating tools for FPS video games, posing a serious threat to the gaming community as well as the entire gaming industry. Targeting this issue, the purpose of our project is to develop a system with robust adversarial algorithms that can be used by the gaming companies to defend against AI cheating in FPS games.<\/td>\n <\/tr>\n
\n MISSION<\/b><\/font><\/td>\n The objective of our project is to:\n1. Create defense and penalty algorithms that can work in real-time video games:\n1a. Defense algorithm needs to add perturbation to the body of the opponent so that object detectors (e.g. Yolov3) will misclassify the person into a different class.\n1b. Penalty algorithm needs to add perturbation to a random background region such that the object detector will detect that region as a person and fool the tool to shoot.\n2. Build a 3D modeling game environment, where players can have the basic operations of free movement and shooting systems. The complete 3D game environment can provide us more freedom to the different kinds of tests we want.<\/td>\n <\/tr>
\n TEAM\n <\/b><\/font><\/td>\n Xicheng Guo
Jintao Cao
Linyuan Zhang
Zilin Chen
Hieu Phan<\/td>\n <\/tr>\n
\n SITE <\/b><\/font><\/td>N\/A<\/td>\n <\/tr>\n<\/tbody>\n<\/table>\n\n\n\n
\n