Abstract: Abstract Most research for computer input consider eye gaze for pointing as an intuitive and promising interaction. The big obstacle for eye gaze as pointing device is the low accuracy. To improve on that, the current study proposed gaze-assisted bubble cursor as a more effective way combing a user’s gaze with bubble cursor. And we carried out an empirical experiment to verify the effect of gaze-assisted bubble cursor (compared to the point cursor and the bubble cursor). In this paper, we reviewed the related work for eye gaze as pointing device and the pointing enhancement technology. Then, we present gaze-assisted bubble cursor, which take full advantage of bubble cursor’s effective width and avoid the visual disturbance, as shown in Fig.2. After that, we described a controlled experiment. The experiment testbed was developed in C# and the interface of experiment as shown in Fig.3. A SMI iViewX-RED eye tracker (0.5° accuracy and 120Hz sampling rate) was used to gather gaze data. To make eye gaze more stable, an adaptive low-pass filtering algorithm was used. The experiment employed a 3×3 mixed model design. The two manipulated variables were as follows: (a) For the cursor type: point cursor, bubble cursor, gaze-assisted bubble cursor; (b) For the target location: close range target in the vertical direction, close range target in the horizontal direction, long range target in the horizontal direction. Sixty right-handed volunteers participated in the study (33 females and 27 males), and randomly assigned to 3 levels of cursor type. For each group, participants were asked to finish 120 times of pointing task as soon as possible on the premise of guaranteeing accuracy of target acquisition, with the accuracy, the task time and the task load index (NASA-TLX) were recorded. The mean and standard deviation under each condition and satisfaction evaluation were shown in Table 1 and Table2. The ANOVA results showed: 1) gaze-assisted bubble cursor outperform the point cursor on the performance (p<.001) and get lower temporal demand 2) gaze-assisted bubble cursor is inferior to bubble cursor in the task time (p<.05) and get higher effort 3) There was significant difference among the different target location, and the target location has the greatest effect on gaze-assisted bubble cursor. Based on the aforementioned results, it can be concluded that gaze-assisted bubble cursor get higher accuracy indeed and get the same level of the bubble cursor which one of most promising techniques. Moreover, gaze is promising to bridge the gap between a user and a distant display which traditional mouse input does not apply well especially for the disabled. In summary, these findings provide a new thought for the future research on optimizing the gaze-assisted pointing technique. Key words gaze-assisted pointing, bubble cursor, pointing augmented, fine positioning

Key words: gaze-assisted pointing, bubble cursor, pointing augmented, fine positioning

摘要： 摘 要 针对视线交互技术中的点击定位问题，本研究在视线辅助点击技术的基础上引入气泡光标技术，提出了视线辅助气泡光标技术，并考察了该技术在精确定位方面的作用。结果显示：1)视线辅助气泡光标的点击操作绩效显著优于手控尖点光标，并且时间知觉更低；2)视线辅助气泡光标的点击操作时间显著慢于手控气泡光标，并且主观努力程度更高；3)目标位置对三种光标类型的操作绩效均有显著影响，且对视线辅助气泡光标的影响作用最大。本研究的结果为视线辅助点击技术的优化提供了新思路。 关键词 视线辅助点击 气泡光标 点击增强技术 精确定位

关键词: 视线辅助点击 气泡光标 点击增强技术 精确定位