A visual perception state detection approach based on implicit semi-Markov modeling is presented to address real-time and accuracy limitations of traditional methods used in student pilot training.From the perspective of visual perception,the paper explores the changes in student pilots'visual gaze when they perform different flight tasks and their gaze rule scanning strategies in different contexts.The proposed model successfully combines the implicit state with temporal correlation,achieving real-time detection of student pilot's visual perception state.The results show that in the same task scenario,this method exhibits an accuracy of 93.55％,increasing by 13.55％compared to the hidden Markov model,and there is a significant improvement in its detection performance.
关键词
视觉感知/模拟训练/状态检测/隐式半马尔可夫模型
Key words
visual perception/simulation training/state detection/hidden semi-Markov model
维普
万方数据