A Case Study of EmoNeXt-Tiny Without Self-Attention for Facial Emotion Recognition
Online First: 24/06/2026
Corressponding author's email:
tnminh.cdktcn@khanhhoa.edu.vnDOI:
https://doi.org/10.54644/jte.2026.2109Keywords:
Facial Emotion Recognition, ConvNeXt-Tiny, Spatial Transformer Network, Squeeze-and-Excitation, Self-Attention, EmoNeXt-TinyAbstract
Facial emotion recognition (FER) on FER2013 remains challenging due to low-resolution grayscale images, class imbalance, and limited data diversity. This study presents a controlled Tiny-scale ablation of EmoNeXt-Tiny to isolate the contribution of the self-attention (SA) regularization term. Specifically, SA regularization is removed from the objective function, while all other components and conditions are preserved, including the ConvNeXt-Tiny backbone, the Spatial Transformer Network (STN), Squeeze-and-Excitation (SE) modules, the preprocessing pipeline, and the training/evaluation protocol. Experiments are conducted on FER2013 using the official train/validation/test split. Under an identical setup, the SA-free variant achieves 72.95% top-1 test accuracy, compared with 73.34% for the full EmoNeXt-Tiny and 71.99% for the ConvNeXt-Tiny baseline. Because the present study reports a single controlled run per configuration (i.e., without multi-seed repetition), the 0.39 percentage-point gap should be interpreted as a preliminary observation that may fall within random variance. Within this constraint, the findings indicate diminishing returns from SA regularization in this Tiny regime once geometric alignment and channel re-weighting are already incorporated. In addition, the SA-free model simplifies optimization by removing an auxiliary loss component and its associated tuning burden. Overall, an STN+SE-enhanced ConvNeXt-Tiny without SA offers a practical accuracy–complexity trade-off for resource-constrained FER applications.
Downloads: 0
References
F. Chollet, “Xception: Deep Learning with Depthwise Separable Convolutions,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2017, pp. 1251–1258. DOI: https://doi.org/10.1109/CVPR.2017.195
K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016, pp. 770–778. DOI: https://doi.org/10.1109/CVPR.2016.90
I. J. Goodfellow, D. Erhan, P. L. Carrier, A. Courville, M. Mirza, B. Hamner, et al., “Challenges in Representation Learning: A Report on Three Machine Learning Contests,” in Proc. Int. Conf. Neural Inf. Process. (ICONIP), 2013, pp. 117–124. (FER2013). DOI: https://doi.org/10.1007/978-3-642-42051-1_16
Z. Liu, H. Mao, C. Y. Wu, C. Feichtenhofer, T. Darrell, and S. Xie, “A ConvNet for the 2020s,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2022, pp. 11976–11986. DOI: https://doi.org/10.1109/CVPR52688.2022.01167
S. Woo, S. Debnath, R. Hu, X. Chen, Z. Liu, I. S. Kweon, and S. Xie, “ConvNeXt V2: Co-Designing and Scaling ConvNets with Masked Autoencoders,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2023, pp. 16133–16143. DOI: https://doi.org/10.1109/CVPR52729.2023.01548
M. Jaderberg, K. Simonyan, A. Zisserman, and K. Kavukcuoglu, “Spatial Transformer Networks,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2015.
J. Hu, L. Shen, and G. Sun, “Squeeze-and-Excitation Networks,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2018, pp. 7132–7141. DOI: https://doi.org/10.1109/CVPR.2018.00745
R. Pecoraro, V. Basile, and V. Bono, “Local Multi-Head Channel Self-Attention for Facial Expression Recognition,” Information, vol. 13, no. 9, p. 419, 2022. DOI: https://doi.org/10.3390/info13090419
A. P. Fard and M. H. Mahoor, “Ad-Corre: Adaptive Correlation-Based Loss for Facial Expression Recognition in the Wild,” IEEE Access, vol. 10, pp. 26756–26768, 2022. DOI: https://doi.org/10.1109/ACCESS.2022.3156598
O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” in Proc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent. (MICCAI), 2015, pp. 234–241. DOI: https://doi.org/10.1007/978-3-319-24574-4_28
S. Vignesh, M. Savithadevi, M. Sridevi, and R. Sridhar, “A Novel Facial Emotion Recognition Model Using Segmentation VGG-19 Architecture,” Int. J. Inf. Technol., 2023, pp. 1–11. DOI: https://doi.org/10.1007/s41870-023-01184-z
Y. El Boudouri and A. Bohi, “EmoNeXt: An Adapted ConvNeXt for Facial Emotion Recognition,” arXiv preprint arXiv:2501.08199, Jan. 14, 2025.
B. Nguyen et al., “An Efficient Real-Time Emotion Detection Using Camera and Facial Landmarks,” in Proc. 7th Int. Conf. Inf. Sci. Technol. (ICIST), 2017, pp. 251–255. DOI: https://doi.org/10.1109/ICIST.2017.7926765
H. Nguyen, N. Tran, H. D. Nguyen, L. Nguyen, and K. Kotani, “KTFEv2: Multimodal Facial Emotion Database and Its Analysis,” IEEE Access, vol. 11, pp. 17811–17822, 2023, doi: 10.1109/ACCESS.2023.3246047. DOI: https://doi.org/10.1109/ACCESS.2023.3246047
C. Szegedy, W. Liu, and Y. Jia, “Going Deeper with Convolutions,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2015, pp. 1–9, doi: 10.1109/CVPR.2015.7298594. DOI: https://doi.org/10.1109/CVPR.2015.7298594
S. Minaee, M. Minaei, and A. Abdolrashidi, “Deep-Emotion: Facial Expression Recognition Using Attentional Convolutional Network,” Sensors, vol. 21, no. 9, Art. no. 3046, 2021. DOI: https://doi.org/10.3390/s21093046
C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the Inception Architecture for Computer Vision,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016, pp. 2818–2826. DOI: https://doi.org/10.1109/CVPR.2016.308
Downloads
Published
How to Cite
Issue
Section
Categories
License
Copyright (c) 2026 Journal of Technical Education Science
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Copyright © JTE.
