抄録
Using sensors, especially visual, to track people and understand their behavior, has become an active research field (particularly in the computer vision community). This challenging task can be made easier (relatively) by using wearable computers. In our proposed research, we concentrate on the fact that a wearable eye-tracker is used. With created some virtual web screens, we detected the confused state of a user from his/her eye movements. The first research found a special eye movement called "scanning". We applied the technique to detect the confused state, while the user asked some unanswerable questions. We finally discriminated the state with at least 70% success rate by using neural networks pattern recognition. Based op this result, we can create useful computer screen, which is easy to understand and easy to use in web-browsing.
