Kołodziej Marcin, Majkowski Andrzej, Rak Remigiusz J, Francuz Piotr, Augustynowicz Paweł
Warsaw University of Technology, Poland.
John Paul II Catholic University of Lublin, Poland.
J Eye Mov Res. 2018 May 24;11(3). doi: 10.16910/jemr.11.3.3.
In this article, we aimed to present a system that enables identifying experts in the field of visual art based on oculographic data. The difference between the two classified groups of tested people concerns formal education. At first, regions of interest (ROI) were determined based on position of fixations on the viewed picture. For each ROI, a set of features (the number of fixations and their durations) was calculated that enabled distinguishing professionals from laymen. The developed system was tested for several dozen of users. We used k-nearest neighbors (k-NN) and support vector machine (SVM) classifiers for classification process. Classification results proved that it is possible to distinguish experts from non-experts.
在本文中，我们旨在提出一种系统，该系统能够基于眼动数据识别视觉艺术领域的专家。两组受试人员之间的差异在于正规教育程度。首先，根据对所观看图片的注视位置确定感兴趣区域（ROI）。对于每个ROI，计算一组特征（注视次数及其持续时间），这些特征能够区分专业人员和非专业人员。所开发的系统对几十名用户进行了测试。我们使用k近邻（k-NN）和支持向量机（SVM）分类器进行分类过程。分类结果证明，区分专家和非专家是可能的。
