The goal of the Bomb in Hand game is to develop a combat game where the player can control their characters through gestures and fight against bots across a deformable landscape. The player must throw bombs towards the opponents and beat them to win the game.
Bomb in Hand game is a strategy game. Each player controls the character to move or throw a bomb by specific gestures.
Both the player and the bot may move around their terrain. During the course of the game, the player can move and use bombs to attack and kill the bot, thereby winning the game. When bombs are thrown, they cause explosions that deform the terrain and cause damage.
The bots are trained by reinforcement learning and clever enough to dodge the incoming bomb. Our bot can adjust its strategy to not be easily beaten by the player in real-time during the game.
The player wins the game by defeating the bot at every level as fast as possible. We have used different techniques of machine learning to detect the gestures in real time, and reinforcement learning to train the bot.
Project website: link
Zikai Cheng, Linkedin
Yijing Xiao, Linkedin
Jingyi Wang, Linkedin
Wilson Huang, Linkedin
Fanzhe Meng, Linkedin
Jianchen Gao, Linkedin
Xuetao Sun, Linkedin