Facial Expression Analysis in E-learning

• Main Authors: Jyun-Lin Wu, 吳俊霖
• Other Authors: Mau-Tsuen Yang
• Format: Others
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/e9drsj

碩士 === 國立東華大學 === 資訊工程學系 === 102 === This thesis proposed a vision-based e-Learning system that analyzed students’ facial expressions to acquire their learning states. Firstly, Active Shape Model was used to align and track facial feature points. Secondly, eye region, brow region, and mouth region were extracted from captured images. The information inside these regions were quantified to form feature vectors. Thirdly, optical flow was utilized to track head shaking and nodding actions. Subsequently, medium-level facial actions were recognized based on low-level image features. Finally, four high-level learning states were estimated using regression models which were trained by manually-marked ground truths. The proposed learning system offers teacher real-time information about students’ learning affects. As a result, teachers can arrange course curriculum and adjust teaching strategy according to students’ learning states to improve their learning interests and promote their learning motivations.