| 研究実績の概要 |
Research Objectives: Many authentication and communication systems nowadays make use of eye tracking. For example, eye tracking can be used to select objects on a display, such as letters or objects, to perform eye typing or to input password information. When using eye tracking, users need to focus their eyes on each visual object for a certain amount of time to select them, which is called “dwell time”. Eye tracking applications so far have used a wide range of dwell times in between 100 to 1100 ms, however, without regard of user preferences and abilities. The objective of this proposal is to identify the best dwell time for individual users, depending on user age and the type of objects that need to be selected. In particular, we want to develop and test a calibration program that solves the “Midas Touch problem”: object selection mistakes during eye tracking that occur due to an inappropriate dwell time.
Research Performance: Psychophysical experiments have been performed to identify users' preferred dwell time using eye tracking. Users were asked to generate a password consisting of 4 or 6 visual objects. These objects were alphanumeric characters (i.e., letters or numbers), a string of dots, or visual icons. Various dwell times were used in between 200 ms to 2000 ms. Object selection time, selection mistakes, and preferred object dwell time were obtained. Overall, the results showed that a dwell time of 600 ms can be considered as appropriate for object selection using eye tracking.
|
| 今後の研究の推進方策 |
Plans for the Research Scheme: The next step in the project is to investigate 3 important aspects with regard to object selection dwell time using eye tracking.
1. We will investigate whether the users' preferred dwell time varies with age. The research so far has mainly employed young adults (university students), and the next step is to compare dwell time preferences for users from various age groups, including middle-aged and elderly persons over 60 years of age. It is likely that preferred dwell times vary depending on the user's ability to fix the eye gaze on a given object and overall object search time. Both these aspects might be influenced by an age-related decrease in eye sight and fixation ability.
2. We will investigate whether the preferred dwell time of 600 ms, as established in our research so far, indeed is suitable for various object types (alphanumeric characters, dot patterns, and visual icons). Familiarity with the objects on the display may influence preferred dwell time. Results seem to suggest that the dwell time for alphanumeric characters (i.e., letters and digits), can be a bit shorter than for dot patterns and visual icons. This needs to be confirmed.
3. Related to 2., we intend to investigate whether practice and learning effects influence preferred dwell time. It is very likely that practicing the task and increased familiarity with the to-be-selected visual objects will result in fewer object selection mistakes and a preference for a shorter dwell time.
We aim to investigate the above 3 research plans will be investigated in 2 experiments.
|
| 次年度使用額が生じた理由 |
An amount of 230,000 JPY has been carried over from the last fiscal year. The reason is mainly that travel expenses were not used, due to the Covid-19 situation. Although we planned to present our results at an academic conference, we have not travelled. We would like to carry-over the budget and allocate it to recruit participants to our experiments if travel expenses cannot be used the following year either.
|
