2019 Fiscal Year Final Research Report
Estimation of Invisible Information Using Motion of Gaze of Opponent in Incomplete Information Game
| Project/Area Number |
18K18172
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 62040:Entertainment and game informatics-related
|
|---|
| Research Institution | Nihon University (2019)
Teikyo University (2018) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2018-04-01 – 2020-03-31
|
| Keywords | 情報推定 / 視線検出 / 麻雀 |
|---|
| Outline of Final Research Achievements |
Each player has different information in incomplete information games such as Mahjong. Guessing invisible information is important to win a game.In this study, I suggest a method to guess other players' Mahjong tiles by capturing their eye movement and gaze points while playing games.
Consequently, we developed online Mahjong application that can record players' gaze while playing.Preliminary experiments show that players' eye movement relates to tiles they have at the time.
In the future, we plan to confirm versatility of the method through more subject experiments.
|
| Free Research Field |
ゲーム情報学
|
| Academic Significance and Societal Importance of the Research Achievements |
オセロやチェスに加え、近年では囲碁や将棋などの選択肢のより多いゲームにおいてもコンピュータが人間を凌駕する時代になっている。しかし麻雀などの、見えない情報があったり、運が絡んだりするようなゲームにおけるコンピュータ開発の研究は多くない。本研究のような研究を重ねることで、見える情報を基にゲームプランを立てるのみでなく、見えない情報を推定し、それを基に戦略を立てるようなシステムの発展につながると考えている。
|
