Advanced Ensemble Deep Random Vector Functional Link for Eye-Tracking-based Situation Awareness Recognition

被引：4
作者：
Li, Ruilin [1 ,2 ]
Gao, Ruobin [3 ]
Cui, Jian [4 ]
Suganthan, P. N. [1 ,5 ]
Sourina, Olga [2 ]
机构：
[1] Nanyang Technol Univ, Sch Elect & Elect Engn, Singapore, Singapore
[2] Fraunhofer Singapore, Singapore, Singapore
[3] Nanyang Technol Univ, Sch Civil & Environm Engn, Singapore, Singapore
[4] Zhejiang Lab, Artificial Intelligence Res Inst, Res Ctr AI Adv Theory, Hangzhou, Zhejiang, Peoples R China
[5] Qatar Univ, Coll Engn, KINDI Ctr Comp Res, Doha, Qatar
来源：
2022 IEEE SYMPOSIUM SERIES ON COMPUTATIONAL INTELLIGENCE (SSCI) | 2022年
关键词：
Conditionally automated driving; situation awareness (SA); eye-tracking (ET); ensemble deep random vector functional link (edRVFL); pruning; weighting;
D O I：
10.1109/SSCI51031.2022.10022019
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
Situation awareness (SA) plays a significant role in takeover transitions from autonomous to manual driving. Previous researchers have shown that eye movement signals can be used for SA recognition. Moreover, ensemble deep random vector functional link (edRVFL) has demonstrated its superiority in different applications. Therefore, this work proposes an advanced edRVFL (AedRVFL) to perform eye-tracking (ET)based SA recognition, including improvements in two aspects. Specifically, pruning in the direct links is implemented to improve the efficiency of linear features. Then, a weighting method based upon the regression errors of samples is developed. The experiment was conducted on a public conditionally automated driving dataset. Results showed that the proposed AedRVFL outperformed the baseline methods, demonstrating the effectiveness of using AedRVFL for ET-based SA recognition. Ablation studies were conducted to validate the improvements in the edRVFL.
引用
收藏
页码：300 / 307
页数：8
相关论文
共 29 条
[1]   Online Recognition of Driver-Activity Based on Visual Scanpath Classification [J].
Braunagel, Christian ;
Geisler, David ;
Rosenstiel, Wolfgang ;
Kasneci, Enkelejda .
IEEE INTELLIGENT TRANSPORTATION SYSTEMS MAGAZINE, 2017, 9 (04) :23-36
[2]   Driver-Activity Recognition in the Context of Conditionally Autonomous Driving [J].
Braunagel, Christian ;
Stolzmann, Wolfgang ;
Kasneci, Enkelejda ;
Rosenstiel, Wolfgang .
2015 IEEE 18TH INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS, 2015, :1652-1657
[3]   Random forests [J].
Breiman, L .
MACHINE LEARNING, 2001, 45 (01) :5-32
[4]   Mapping Brain Activity During Loss of Situation Awareness: An EEG Investigation of a Basis for Top-Down Influence on Perception [J].
Catherwood, Di ;
Edgar, Graham K. ;
Nikolla, Dritan ;
Alford, Chris ;
Brookes, David ;
Baker, Steven ;
White, Sarah .
HUMAN FACTORS, 2014, 56 (08) :1428-1452
[5]   XGBoost: A Scalable Tree Boosting System [J].
Chen, Tianqi ;
Guestrin, Carlos .
KDD'16: PROCEEDINGS OF THE 22ND ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2016, :785-794
[6]   Why do car drivers fail to give way to motorcycles at t-junctions? [J].
Crundall, David ;
Crundall, Elizabeth ;
Clarke, David ;
Shahar, Amit .
ACCIDENT ANALYSIS AND PREVENTION, 2012, 44 (01) :88-96
[7]   Situation awareness based on eye movements in relation to the task environment [J].
de Winter, J. C. F. ;
Eisma, Y. B. ;
Cabrall, C. D. D. ;
Hancock, P. A. ;
Stanton, N. A. .
COGNITION TECHNOLOGY & WORK, 2019, 21 (01) :99-111
[8]   Examining the effects of emotional valence and arousal on takeover performance in conditionally automated driving [J].
Du, Na ;
Zhou, Feng ;
Pulver, Elizabeth M. ;
Tilbury, Dawn M. ;
Robert, Lionel P. ;
Pradhan, Anuj K. ;
Yang, X. Jessie .
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, 2020, 112 :78-87
[9]  
Durso F. T., 1998, Air Traffic Control Q, V6, P1, DOI [https://doi.org/10.2514/atcq.6.1.1, DOI 10.2514/ATCQ.6.1.1, 10.2514/atcq.6.1.1]
[10]   TOWARD A THEORY OF SITUATION AWARENESS IN DYNAMIC-SYSTEMS [J].
ENDSLEY, MR .
HUMAN FACTORS, 1995, 37 (01) :32-64
123