Deep Learning-Based LecturePosture Evaluation
基于深度学习的讲课姿态评估作者机构:Key Laboratory of Multidisciplinary Management and Control of Complex Systems of Anhui Higher Education InstitutesAnhui University of TechnologyMaanshan 243002AnhuiChina School of Microelectronics and Data ScienceAnhui University of TechnologyMaanshan 243032AnhuiChina
出 版 物:《Wuhan University Journal of Natural Sciences》 (武汉大学学报(自然科学英文版))
年 卷 期:2024年第29卷第4期
页 面:315-322页
核心收录:
学科分类:0810[工学-信息与通信工程] 08[工学] 081002[工学-信号与信息处理]
基 金:Supported by the Open Fund of Key Laboratory of Anhui Higher Education Institutes(CS2021-07) the National Natural Science Foundation of China(61701004) the Outstanding Young Talents Support Program of Anhui Province(gxyq2021178)
主 题:deep learning human pose estimation object detection correlation
摘 要:Computer vision,a scientific discipline enables machines to perceive visual information,aims to supplant human eyes in tasksencompassing object recognition,localization,and *** traditional educational settings,instructors or evaluators evaluate teachingperformance based on subjective ***,with the continuous advancements in computer vision technology,it becomes increasinglycrucial for computers to take on the role of judges in obtaining vital information and making unbiased *** thisbackdrop,this paper proposes a deep learning-based approach for evaluating lecture ***,feature information is extracted fromvarious dimensions,including head position,hand gestures,and body posture,using a human pose estimation ***,a machinelearning-based regression model is employed to predict machine scores by comparing the extracted features with expert-assigned *** correlation between machine scores and human scores is investigated through experiment and analysis,revealing a robustoverall correlation(0.6420)between predicted machine scores and human *** ideal scoring conditions(100 points),approximately51.72%of predicted machine scores exhibited deviations within a range of 10 points,while around 81.87%displayed deviationswithin a range of 20 points;only a minimal percentage of 0.12%demonstrated deviations exceeding the threshold of 50 ***,tofurther optimize performance,additional features related to bodily movements are extracted by introducing facial expression recognitionand gesture recognition *** fusion of multiple models resulted in an overall average correlation improvement of 0.0226.