Application of Xinghuo Large Model in Teachers’ Facial Emotion Analysis
Journal of Contemporary Educational Research · 07.04.2026 · 10 (3) · 239-244
DOI: 10.26689/jcer.v10i3.14430 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Contemporary Educational Research, online erschienen am 07.04.2026. Verfasst von Kai Xu und 1 weiteren Autor:innen.
Abstract (Originalsprache)
Facial expression recognition has been an active research field over the past few decades, with typical methods including principal component analysis based on eigenfaces and independent component analysis. With the development of deep learning technology, convolutional neural networks have also played an important role in facial expression recognition. Although these methods perform well, there is still significant room for improvement. This paper uses the Xinghuo Large Model for teachers’ facial emotion analysis. First, classroom recorded videos are framed to extract key facial expression regions of teachers; then a network model is constructed, which is adjusted and extracted through a two-stream architecture; next, teachers’ facial data are used as network input, and emotion prediction is performed using deep learning-based methods and Xinghuo Large Model-based methods respectively; finally, the prediction results are fused to obtain the final teachers’ facial emotion analysis results.
Zitieren
APA
Xu, K., & Tao, L. (2026). Application of Xinghuo Large Model in Teachers’ Facial Emotion Analysis. Journal of Contemporary Educational Research, 10(3), 239-244. https://doi.org/10.26689/jcer.v10i3.14430
Vancouver
Xu K, Tao L. Application of Xinghuo Large Model in Teachers’ Facial Emotion Analysis. Journal of Contemporary Educational Research. 2026;10(3):239-244. doi:10.26689/jcer.v10i3.14430
BibTeX
@article{Xu2026,
author = {Xu, Kai and Tao, Lang},
title = {Application of Xinghuo Large Model in Teachers’ Facial Emotion Analysis},
journal = {Journal of Contemporary Educational Research},
year = {2026},
volume = {10},
number = {3},
pages = {239--244},
doi = {10.26689/jcer.v10i3.14430}
}
EUVIMED