Don't Panic: A VR Zombie Shooter

2019. Virtual Reality. Machine Learning.

The objective of this research was to experiment with different ways of obtaining biofeedback through virtual reality (VR) headsets and use them to somehow modify the user experience to make each gameplay unique and interesting for the player. Our experiment focuses on the feedback that can be obtained from the hand movements of the player and attempts to adjust the difficulty of the game based on it. It is known that user-inputs (biofeedback) can be obtained through most standard VR headsets, which is why the experiment pays more emphasis on merging VR with Artificial Intelligence (AI). The AI helps analyze the user-inputs and classify them to build relationships between the virtual world and the movements of the player.

The game we developed consists of a “Zombie-Shooter” type game in which the player must survive waves of incoming zombies. The difficulty of the game increases based on the arousal (excited/nervousness) of the player as well as the progression of the game. The arousal of the player is determined by the AI through the analysis of hand movements. By doing this experiment we hope to better understand the interconnectivity between these two emerging fields of technology i.e. Virtual Reality and Artificial Intelligence.