基于视线跟踪的智能界面实现机制研究

ON IMPLEMENTATION MECHANISM OF INTELLIGENT INTERFACE BASED ON GAZE TRACKING

针对传统用户界面难以满足普适计算发展的需求,提出一种基于视线跟踪的智能界面解决方案。首先,以视觉注意力选择机制为理论基础,对智能界面交互框架中用户、界面和系统三要素进行统一描述和说明;然后,以三要素的映射关系为研究依据,采用模糊聚类方法来计算用户对界面对象的注意程度值,并使用视觉注意分配策略来预测用户对界面对象的兴趣程度;最后,设计相关实验来验证智能界面实现机制的有效性。实验结果表明,智能界面预测用户决策意图的准确率高达85%,是一种能够主动标注用户兴趣目标的智能界面可选方案。

Because traditional user interfaces are difficult to meet the requirement of the development of pervasive computing, we proposed a gaze tracking-based intelligent interface solution. First, we took the mechanism of visual attention selection as the theoretical basis, and gave the unified description and interpretation on three key factors of user, interface and system in interaction framework of intelligent interface. Next, according to mapping relationship among three key factors, we used fuzzy clustering method to compute the level of users＇ attention on interface objects, and used visual attention allocation strategy to predict interest degree of users on interface objects. Finally, we designed the related experiments to verify the effectiveness of implementation mechanism of the proposed intelligent interface. Experimental results indicated that the proposed intelligent interface reached the accuracy up to 85% in predicting users＇ intention of decision making, and was an optional solution that could actively mark users＇ interesting objects.