Author(s)

JuanYang¹, Ling Ma^2*, XianBin Zhang², SiYi Tian¹

Affiliation(s)

¹School of Journalism and Law, Wuchang Shouyi University, Wuhan 430064, Hubei, China.

²School of Information Science and Engineering, Wuchang Shouyi University, Wuhan 430064, Hubei, China.

Corresponding Author

Ling Ma

ABSTRACT

Students’ facial micro-expressions can reflect their learning states, and the intelligent recognition of micro-expressions is of great significance for monitoring these states. This paper proposes a micro-expression recognition method based on a Residual Network for detecting the learning and psychological states of college students. Using ResNet18 as the backbone network, the Efficient Channel Attention (ECA) mechanism is embedded to adaptively adjust channel weights, enhancing feature representation capability. Simultaneously, the Mish activation function and Dropout layer are introduced to optimize gradient flow and reduce the risk of overfitting. During training, the Label Smoothing Cross Entropy loss function, the AdamW optimizer combined with the OneCycleLR learning rate scheduler, and an early stopping mechanism are adopted, effectively improving the model’s generalization ability and training efficiency on small datasets. Experiments are based on a self-built dataset (including positive, neutral, and negative expressions, totaling 1800 grayscale images). Through data augmentation and ten-fold cross-validation, the model achieves an accuracy of 97.50%. The experimental results show that this method possesses high accuracy and robustness in micro-expression recognition tasks, providing an effective tool for monitoring the psychological states of college students and optimizing classroom teaching.

KEYWORDS

Deep learning; Convolutional neural network; Micro-expression recognition; Learning state monitoring

CITE THIS PAPER

JuanYang, Ling Ma, XianBin Zhang, SiYi Tian. Intelligent recognition of students' learning states through micro-expressions. World Journal of Information Technology. 2025, 3(5): 1-6. DOI: https://doi.org/10.61784/wjit3060.

REFERENCES

[1] Tonguc G, Ozkara BO. Automatic recognition of student emotions from facial expressions during a lecture. Computers & Education, 2020(148): 1-12.

[2] Wei YT, Lei F, Hu MJ, et al. Review of Research on Student Expression Recognition. The Chinese Journal of ICT in Education, 2020(21): 48-55.

[3] Ekman P. Facial expression and emotion. American Psychologist, 1993, 48: 384.

[4] Tong XY, Sun SL, Fu MX. Data augmentation and second-order pooling for facial expression recognition. IEEE Access, 2019, 7: 86821-86828.

[5] Huang XL, Gou XS, Chen X. Facial Micro-expression Recognition Algorithm Based on Hybrid Features and Information Entropy. Computer Simulation, 2023, 40(06): 197-201.

[6] Qiao GF, Hou SM, Liu YY. Facial expression recognition algorithm based on improved convolutional neural network and support vector machine. Journal of Computer Applications, 2022, 42(04): 1253-1259.

[7] Chen T, Xing S, Yang WW, et al. Facial expression recognition integrating spatiotemporal features. Journal of Image and Graphics, 2022, 27(07): 2185-2198.

[8] Yang HY, Ciftci U, Yin LJ. Facial Expression Recognition by De-expression Residue Learning. Conference on Computer Vision and Pattern Recognition. IEEE, 2018: 2168-2177.

[9] He KM, Zhang XY, Ren SQ, et al. Deep Residual Learning for Image Recognition Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2016: 770-778.

[10] Wang QL, Wu BG, Zhu PF, et al. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, 2020: 11531-11539. DOI: 10.1109/CVPR42600.2020.01155.

[11] Misra D. Mish: A Self Regularized Non-monotonic Activation Function. 2019: 8. https://arxiv.org/abs/1908.08681.

[12] Hu J, Shen L, Sun G. Squeeze-and-Excitation Networks Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2018: 7132-7141.

[13] He JB, Zhou JX, Gan JH, et al. Classroom Expression Classification Model Based on Multi-task Learning. Journal of Applied Sciences, 2024, 42(06): 947-961.