摘要
虛擬教室結合頭戴式顯示器的系統是目前認知復健上相當有前景的一項虛擬實境發展與應用，由於學習的成效好壞和個人的注意力集中與否有相當大的關係。因此藉由與虛擬教室中人物的互動學習，能夠將注意力不集中的行為矯正回來。注意力好的學習者會將大部分的注意力集中於老師或是白板、投影幕的教材上，反之，注意力差的學習者容易被課堂上其他學生的舉動所影響，導致其無法集中注意力於學習上。為了能夠在虛擬教室系統中了解受試者的注意力所在，本研究透過3D頭戴式顯示器做以科技輔助來為受試者呈現出虛擬實境的教室場景 ，有鑑於眼動儀兼頭戴式顯示器的軟硬體系統價格昂貴，一般教育研究者無法負擔。為了推廣虛擬教室系統，因此本研究即發展出一種以穿戴式感測器資料為基礎之注意力偵測方法來克服此問題，利用了創新的注意力圈設計方法來判斷受試者在虛擬教室實驗中所專注的目標。本研究共徵募30位身體健康的成人進行虛擬教室注意力實驗，實驗過程中每位受試者皆確實聽從研究者實驗的指示並做出反應。實驗結果從頭戴式顯示器所偵測到的頭部轉動角度資訊與注意力圈資訊做事後分析，也利用了SVM的分析方法改善系統分析的準確率，實驗結果證明了注意力圈有很好的準確率並透過SVM的分類分析有助於提高注意力圈偵測的準確率。
並列摘要
The Virtual Classroom (VC) with the head-mounted display (HMD) is a quietly promising development and application in virtual reality for the cognitive rehabilitation nowadays. Since the effect of learning has an important relationship with personal attention focus, using the interactive learning with virtual characters in the VC enables to redress the behavior of patients who have attention deficit. Good learners can pay attention and focus on a teacher, a whiteboard or a projection screen, however, learners with attention deficit are easy to be influenced by other students’ movement and behavior, so that they cannot concentrate on learning. Therefore, understanding what students look at is a very important thing. In order to find out the targets students care, in this study, a virtual classroom system through HMD has an innovative method called the attention ring that can detect their attention. Due to a high price of commercially available HMD with eye-tracking, this study designs an attention detection method with a wearable sensor to overcome the attention detection problem, so that the VC system can be popularized for general education researchers. This study collects 30 healthy participants to attend the virtual classroom attention experiment. During the experiment, every participant listens to the researcher’s instructions to look at the specific targets in the VC. The VC system gives two outputs about the data of the attention ring and the degrees of head movement rotation for post-hoc analysis. As a result, the accuracy of attention detection leads to upon 80%, and using the SVM analysis to improve overall accuracy is verified.
參考文獻
延伸閱讀
- 顏志育（2007）。虛擬內視鏡基礎技術訓練系統學習轉移之研究〔碩士論文，國立中央大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0031-0207200917351017
- Hsieh, W. T. (2022). 用於虛擬實境顯示器之感知解析度提升之高效神經網路及硬體實作 [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU202200629
- 陳又菁（2020）。導入頭戴式虛擬實境於學習成效、心流體驗與認知負荷之探究：以昆蟲課程為例。數位學習科技期刊，12(3)，1-23。https://doi.org/10.3966/2071260X2020071203001
- 陳世宏（2020）。結合虛擬實境與科普教育之遊戲式學習系統開發與評估〔碩士論文，國立宜蘭大學〕。華藝線上圖書館。https://doi.org/10.6820/niu202000106
- 王錫恩（2016）。結合行動感測器與虛擬場景模擬電影運鏡之教學輔助工具〔碩士論文，國立臺中科技大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0061-2701201617272500