With the use of AI we try to detect which of the players in the popular game Among Us, is an Impostor. OCR filters out text from recorded game sessions. The messages are stored in JSON files an for each character, the color, name and message is saved. Subsequently, natural language processing techniques are used to be able to interpret the messages. Next, machine learning techniques such as SVM are used to classify the messages and detect impostors.

The paper is published at the 3rd IEEE Conference on Games.

Team Among Us

Harro, Martin