Abstract: To improve the accuracy and efficiency of eye tracking on mobile device and to reduce the cost,a gaze recall method as a complementation for eye movement data computing on mobile device was proposed.In this method,crowdsourcing technique was applied to distribute gaze recall tasks in crowd sensing environment,and task recommendation algorithm based on context-aware service was used to improve the gaze recall accuracy.A prototype system for gaze recall on mobile device was implemented,and the result of user testing showed that the recalled gaze data were very similar to gaze data from real eye tracker,which could validate the feasibility and effect of the proposed method.

Key words: eye tracking;gaze recall, visual attention, crowdsourcing, context-aware service, human-computer interaction

摘要： 为了提高移动设备上的眼动跟踪精度和效率、降低硬件成本,提出基于注视点回忆的眼动数据计算方法来代替移动设备上的眼动跟踪。应用众包技术在群智环境下实现用户注视点回忆任务的发布,基于上下文感知服务技术进行注视点回忆任务的推荐,以提高回忆注视点精度。面向广告视觉评估设计开发了注视点回忆众包应用系统。用户测试结果表明,回忆注视点数据与眼动仪获取的真实眼动数据有较高的一致性,验证了该方法的可行性和有效性。

关键词: 眼动跟踪, 注视点回忆, 视觉注意, 众包, 上下文感知服务, 人机交互