An extensive study of user identification via eye movements across multiple datasets
作者：
Highlights：
• An extensive study on biometrics via eye movements with many datasets is presented.
• Several factors (gender, age etc.) affecting the eye movement biometrics are studied.
• Three methods to improve the state of the art accuracy are presented.
• State of the art accuracy is improved from 86% to 96%.
• This research affirms the viability of biometrics with off-the-shelf eye trackers.
摘要
•An extensive study on biometrics via eye movements with many datasets is presented.•Several factors (gender, age etc.) affecting the eye movement biometrics are studied.•Three methods to improve the state of the art accuracy are presented.•State of the art accuracy is improved from 86% to 96%.•This research affirms the viability of biometrics with off-the-shelf eye trackers.
论文关键词：Eye tracking,Eye movement biometrics,User identification,Machine learning,IVT algorithm
论文评审过程：Received 26 October 2021, Revised 30 March 2022, Accepted 21 June 2022, Available online 4 July 2022, Version of Record 14 July 2022.
论文官网地址：https://doi.org/10.1016/j.image.2022.116804
