孫弘岳Suen, Hung-Yue劉毓翔Liu, Yu Hsiang2023-12-082028-07-092023-12-082023https://etds.lib.ntnu.edu.tw/thesis/detail/6e9e421067b7ac983a72fe645dc2f41c/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/120464全球冠狀病毒COVID-19疫情大流行,促進教育訓練產業蓬勃發展,在國外知名研究機構的預測中,教育科技產業預算支出將於2025年上看4,000億美元。這樣衝擊實體學習環境的改變將可能不是暫時性的。我們自然更需要有效且快速的工具,去驗證與提醒講師在數位時代的線上學習中,如何持續吸引學習者的注意力,並進一步的促進教學品質有效提升。過去研究顯示,講師的面部情緒表達與預測學員注意力的情感投入習習相關。有鑒於此,本研究以深度學習為工具,粹取講師面部情緒變化,並探討該面部情緒在線上學習的情境中和學員情感投入的相關性。本研究發現在非同步線上學習環境中,學生的情感投入受聲音魅力正向影響,與講師面部恐懼情緒正相關,其他面部情緒對學生情感投入則無顯著影響。本研究還揭示了講師的聲音魅力和面部恐懼情緒對情感投入的影響的同時,還提供了與過去研究觀點不同的發現,有助於推動相關研究領域的發展。並期望做為日後學校、教育機構與企業在執行與設計課程時,提醒講師如何優化進而達到提升訓練效果之論述依據,並幫助講師創造出更有吸引力和有助於學習的教學環境。The COVID-19 pandemic has fueled the growth of the education and training industry, with projected expenditures of $400 billion in educational technology by 2025. The shift towards online learning may be long-lasting, necessitating effective tools to enhance instructor engagement and teaching quality. Prior research indicates a connection between instructors' facial expressions and learner attention. This study employs deep learning techniques to analyze facial emotions of instructors and their correlation with learner engagement in online contexts. The findings affirm the importance of emotional engagement for learning outcomes. In asynchronous online learning, learners' emotional engagement is positively influenced by vocal charm and facial fear expression of instructors. Other facial emotions have no significant impact. These insights contributeto the advancement of research and offer guidance for schools and institutions to optimize training effectiveness by providing real-time tools for instructors.表情外表情緒情感運算深度學習情感投入expressionfacial expressionaffective computingdeep learningemotional engagement線上學習AI模型初探:講師表情與學員情感投入之相關性研究Exploring AI models through online learning: A study on the correlation between instructor expressions and student emotional engagementetd