Research Article
Face Emotion Expression Recognition Using DLIB Model and Convolutional Neural Network Approach for Supporting Online Learning
@INPROCEEDINGS{10.1007/978-3-031-31275-5_15, author={Rita Wiryasaputra and Chin-Yin Huang and Jimmy Juliansyah and Chao-Tung Yang}, title={Face Emotion Expression Recognition Using DLIB Model and Convolutional Neural Network Approach for Supporting Online Learning}, proceedings={Smart Grid and Internet of Things. 6th EAI International Conference, SGIoT 2022, TaiChung, Taiwan, November 19-20, 2022, Proceedings}, proceedings_a={SGIOT}, year={2023}, month={5}, keywords={Convolutional Neural Network (CNN) DLIB Deep Learning Face Emotion Expression Recognition Machine Learning}, doi={10.1007/978-3-031-31275-5_15} }- Rita Wiryasaputra
Chin-Yin Huang
Jimmy Juliansyah
Chao-Tung Yang
Year: 2023
Face Emotion Expression Recognition Using DLIB Model and Convolutional Neural Network Approach for Supporting Online Learning
SGIOT
Springer
DOI: 10.1007/978-3-031-31275-5_15
Abstract
Many sectors have experienced the impact of the COVID-19 outbreak, without exception education. The method of process learning transformed from face-to-face meeting learning became online learning. Learners tried to adapt to this unexpected circumstance. In the online learning approach, the instructors only assumed the degree of learners’ understanding with their face emotion expressions spontaneously. Advancement technology enables the machine to learn data fast and accurately. Mostly, the position of the learner’s face in front of the camera when attending the online course, and the DLIB’s shape detector model map the landmark of the captured face. Deep learning is a subset domain of machine learning. Convolutional Neural Network (CNN) model as a deep learning approach has characteristics in the high computation and ease of implementation. The work proposed a face-emotion expression recognition model for supporting online learning. The combination ratio images dataset was 80% data training and 20% data testing, and the condition expression was determined with a deep learning approach. The experimental results showed that the recognition accuracy of the proposed model achieved 97% for dataset image input.
