Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, BC, Canada V6T 1Z4.
Behav Res Methods. 2011 Sep;43(3):843-52. doi: 10.3758/s13428-011-0078-8.
Given the prevalence, quality, and low cost of web cameras, along with the remarkable human sensitivity to gaze, we examined the accuracy of eye tracking using only a web camera. Participants were shown webcamera recordings of a person's eyes moving 1°, 2°, or 3° of visual angle in one of eight radial directions (north, northeast, east, southeast, etc.), or no eye movement occurred at all. Observers judged whether an eye movement was made and, if so, its direction. Our findings demonstrate that for all saccades of any size or direction, observers can detect and discriminate eye movements significantly better than chance. Critically, the larger the saccade, the better the judgments, so that for eye movements of 3°, people can tell whether an eye movement occurred, and where it was going, at about 90% or better. This simple methodology of using a web camera and looking for eye movements offers researchers a simple, reliable, and cost-effective research tool that can be applied effectively both in studies where it is important that participants maintain central fixation (e.g., covert attention investigations) and in those where they are free or required to move their eyes (e.g., visual search).
鉴于网络摄像头的普及、质量和低成本，以及人类对注视的高度敏感性，我们仅使用网络摄像头检查了眼动追踪的准确性。参与者观看了一个人眼睛以 1°、2°或 3°的视角在八个径向方向（北、东北、东、东南等）之一移动的网络摄像头记录，或者根本没有眼睛移动。观察者判断是否进行了眼动，如果进行了，判断其方向。我们的研究结果表明，对于所有大小或方向的扫视，观察者都可以显著优于随机水平来检测和区分眼动。至关重要的是，扫视越大，判断越好，因此对于 3°的眼动，人们可以在 90%或更高的准确率上判断是否发生了眼动，以及它的去向。这种使用网络摄像头寻找眼动的简单方法为研究人员提供了一种简单、可靠且经济高效的研究工具，无论是在参与者需要保持中心注视的研究中（例如，隐蔽注意力研究），还是在参与者可以自由或需要移动眼睛的研究中（例如，视觉搜索），都可以有效地应用。
