Bridging Accuracy and Efficiency in Facial Expression Recognition: An Adaptive Swarm Intelligence Framework for Resource-Constrained Environments

Author Details

Facial expression recognition (FER) systems confront major computational hurdles due to high-dimensional feature spaces, limiting their application on resource-constrained devices. This research offers a unique Particle Swarm Optimization with Adaptive Opposition-Based Learning (PSO-AOBL) framework for optimal feature selection in FER systems. The suggested adaptive technique prevents premature convergence while preserving quick optimization performance by dynamically initiating opposition-based learning based on swarm diversity. Three benchmark datasets—CK+, FER2013, and JAFFE—are used to assess our methodology. Experimental results show PSO-AOBL reduces features by 73% (from 3,918 to 1,050 features) and inference time by 75% while achieving 94.2% accuracy on the CK+ dataset. Compared to normal PSO, our technique offers 2.1% accuracy improvement and 43% faster convergence.

Keywords

Facial expression recognition, feature selection, particle swarm optimization, swarm intelligence, computational efficiency.

[1] S. Li and W. Deng, “Deep facial expression recognition: A survey,” IEEE Trans. Affective Comput., vol. 13, no. 3, pp. 1195-1215, 2022.

[2] A. Mollahosseini et al., “AffectNet: A database for facial expression,” IEEE Trans. Affective Comput., vol. 10, no. 1, pp. 18-31, 2019.

[3] Y. Zhang et al., “Relative uncertainty learning for facial expression recognition,” Proc. NeurIPS, pp. 17616-17627, 2021.

[4] M. Ghosh et al., “Feature selection for facial expression recognition: A review,” Expert Syst. Appl., vol. 189, article 116077, 2022.

[5] I. Guyon and A. Elisseeff, “An introduction to variable and feature selection,” J. Mach. Learn. Res., vol. 3, pp. 1157-1182, 2003.

[6] G. Chandrashekar and F. Sahin, “A survey on feature selection methods,” Computers Elect. Eng., vol. 40, no. 1, pp. 16-28, 2014.

[7] L. Yu and H. Liu, “Efficient feature selection via analysis of relevance and redundancy,” J. Mach. Learn. Res., vol. 5, pp. 1205-1224, 2004.

[8] M. Dash and H. Liu, “Feature selection for classification,” Intell. Data Anal., vol. 1, no. 3, pp. 131-156, 1997.

[9] M. Dorigo et al., “Ant colony optimization,” IEEE Comput. Intell. Mag., vol. 1, no. 4, pp. 28-39, 2006.

[10] R. Poli et al., “Particle swarm optimization: An overview,” Swarm Intell., vol. 1, no. 1, pp. 33-57, 2007.

[11] J. Kennedy and R. Eberhart, “Particle swarm optimization,” Proc. IEEE ICNN, pp. 1942-1948, 1995.

[12] J. Kennedy and R. C. Eberhart, “A discrete binary version of PSO,” Proc. IEEE SMC, pp. 4104-4108, 1997.

[13] B. Xue et al., “PSO for feature selection: A multi-objective approach,” IEEE Trans. Cybern., vol. 43, no. 6, pp. 1656-1671, 2013.

[14] X. Jiang et al., “DFEW: A large-scale database for dynamic facial expressions,” Proc. ACM MM, pp. 2881-2889, 2020.

[15] M. M. Kabir et al., “A new hybrid ant colony optimization algorithm,” Expert Syst. Appl., vol. 39, no. 3, pp. 3747-3763, 2012.

[16] H. R. Tizhoosh, “Opposition-based learning: A new scheme,” Proc. CIMCA, pp. 695-701, 2005.

[17] P. Ekman and W. V. Friesen, “Facial action coding system,” Consulting Psychologists Press, 1978.

[18] T. Ojala et al., “Local binary patterns,” IEEE Trans. PAMI, vol. 24, no. 7, pp. 971-987, 2002.

[19] N. Dalal and B. Triggs, “Histograms of oriented gradients,” Proc. IEEE CVPR, pp. 886-893, 2005.

[20] K. Wang et al., “Suppressing uncertainties for large-scale FER,” Proc. IEEE CVPR, pp. 6897-6906, 2020.

[21] F. Zhang et al., “Joint pose and expression modeling,” Proc. IEEE CVPR, pp. 3359-3368, 2022.

[22] Y. Li et al., “Self-supervised representation learning,” Proc. IEEE CVPR, pp. 3633-3642, 2021.

[23] K. Kira and L. A. Rendell, “A practical approach to feature selection,” Proc. ICML, pp. 249-256, 1992.

[24] I. Guyon et al., “Gene selection using support vector machines,” Mach. Learn., vol. 46, pp. 389-422, 2002.

[25] S. Ghosh et al., “Automatic group affect analysis,” IEEE Trans. Affective Comput., vol. 14, no. 2, pp. 1169-1180, 2023.

[26] A. Kumar et al., “Feature selection using regularization methods,” Pattern Recognit. Lett., vol. 158, pp. 123-130, 2023.

[27] M. Dorigo and L. M. Gambardella, “Ant colony system,” IEEE Trans. Evol. Comput., vol. 1, no. 1, pp. 53-66, 1997.

[28] B. Xue et al., “A survey on evolutionary computation for feature selection,” IEEE Trans. Evol. Comput., vol. 20, no. 4, pp. 606-626, 2016.

[29] J. Too et al., “Hybrid binary PSO differential evolution,” IEEE Access, vol. 8, pp. 140753-140769, 2020.

[30] M. Mafarja and S. Mirjalili, “Hybrid whale optimization algorithm,” Neurocomputing, vol. 260, pp. 302-312, 2017.

[31] S. Rahnamayan et al., “Opposition-based differential evolution,” IEEE Trans. Evol. Comput., vol. 12, no. 1, pp. 64-79, 2008.

[32] H. Wang et al., “Enhancing PSO using generalized opposition-based learning,” Inf. Sci., vol. 181, no. 20, pp. 4699-4714, 2011.

[33] Q. Xu et al., “A review of opposition-based learning,” Eng. Appl. Artif. Intell., vol. 29, pp. 1-12, 2014.

[34] S. Ergezer et al., “Oppositional biogeography-based optimization,” Proc. IEEE SMC, pp. 1009-1014, 2009.

[35] D. Patel et al., “PSO for feature selection in FER,” Proc. IEEE ICIP, pp. 2145-2150, 2024.

[36] Z. Zeng et al., “A survey of affect recognition methods,” IEEE Trans. PAMI, vol. 31, no. 1, pp. 39-58, 2009.

[37] V. Kazemi and J. Sullivan, “One millisecond face alignment,” Proc. IEEE CVPR, pp.

Sharma and A. Kumar, “Bridging accuracy and efficiency in facial expression recognition: An adaptive swarm intelligence framework for resource-constrained environments,” IPEM Journal of Computer Application & Research, vol. 10, pp. 13–19, Dec. 2025, doi: