Research on Remote Intelligent Monitoring System for Online Examination Based on Gaze Direction Classification

• DOI: 10.1145/3594315.3594369
• Publication Date: 2023-08-03T00:14:16Z
• Authors: Zuhui Hu; Yaguang Jing; Guoqing Wu
• Keywords: Recall rate; Deep belief network

Aiming at the problem that online examination is difficult to monitor, a remote intelligent monitoring system for based on gaze direction classification proposed. In experimental environment, situation, dataset with nine types of directions multi-classification problems established, and image model used training. The same comparative experiments ConvNeXt network ResNet network, finally selected as backbone network. trained performs well in weighted accuracy recall rate test set, reaching 88.82% 86.96%. results show solution feasible. It can not only improve mode, but also have low economic cost, which great significance promote popularization examination.