Portal de Periódicos da CAPES

A Brain-Controlled Mahjong Game with Artificial Intelligence Augmentation

2022; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-031-20503-3_47

1611-3349

Xiaodi Wu, Yu Qi, Xinyun Zhu, Kedi Xu, Junming Zhu, Jianmin Zhang, Yueming Wang,

Advanced Memory and Neural Computing

Brain-computer interface (BCI) provides a direct pathway from the brain to external devices, which have demonstrated potential in rehabilitation and human-computer interaction. One limitation of current BCI lies in the low precision and stability in control, thus the user scenario and the interaction process play an important role in BCI applications. We propose to leverage the ability of artificial intelligence (AI) to improve the easy-of-use of BCIs. We designed a brain-controlled mahjong game with AI augmentation. As the user plays the game, an AI system continuously monitors the game and suggests several optimal candidate options, and the user directly selects from the candidates rather than from all the mahjong tiles in hand. This design enables the mahjong game with easy control, thus improving the user experience. Our system is evaluated by a human subject with an invasive BCI system. Results demonstrate that the subject can fluently control the system to play the mahjong game.