A Method of Attention Analysis on Video. Issue 1 (1st April 2022)

• Record Type: Journal Article
• Authors: Li, Wanping; Wu, Jiajie; Yin, Kuiying; Jiang, Guang; Yu, Chao; Li, Lanyu
• Is Part Of: Journal of physics. Volume 2253:Issue 1(2022)
• Publication Date: 2022-04-01
• Subjects: infrared camera -- depth camera -- temporary shift module (TSM) -- attention analysis -- CNN; Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/2253/1/012032
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Attention monitoring system is important for various tasks such as driving by alarming the person when he or she is not attending to the task at hand. Past research has not explored a usable attention monitoring system. In the current study, we used eye trackers, depth camera, and infrared cameras to assess the attention of the participants as they read texts. We extracted features from eye tracking and camera data, and then used convolutional neural network to predict the attention state of the participants. We found the eye tracker data yielded a 90% accuracy in predicting attentional state of the subjects. The camera data yielded over 70% accuracy in prediction.