A Comparative Study of K-Means, Gaussian Mixture Model, and Spectral Clustering for Facial Emotion Recognition

Francesco Adhimas Totti, Nina Setiyawati

Abstract


Facial emotion recognition is one of the key areas of artificial intelligence that has been widely applied in human–computer interaction. This study compares three clustering algorithms—K-Means, Gaussian Mixture Model (GMM), and Spectral Clustering—to group facial expressions from the FER-2013 dataset, which was sampled down to 1,820 images. Facial features were extracted using Gabor filters and reduced in dimensionality using Principal Component Analysis (PCA) to efficiently preserve essential information. The evaluation was conducted using Silhouette Score, Davies–Bouldin Index, and clustering accuracy estimation. The results show that Spectral Clustering achieved the best performance, with a Davies–Bouldin Index of 0.464 and an accuracy of 95.36%, followed by GMM (Silhouette Score 0.526) and K-Means (Silhouette Score 0.523). Furthermore, PCA with 80% variance retention produced an effective 1D feature representation, allowing clustering results to be visualized in a simple yet informative manner. These findings suggest that the choice of clustering algorithm should be aligned with the desired trade-off between system accuracy and efficiency.

Keywords


facial emotion recognition; clustering; gabor filter, PCA

Full Text:

PDF

References


B. R. Delazeri, L. L. Vera, J. P. Barddal, A. L. Koerich, and S. B. Alceu De, “Evaluation of Self-Taught Learning-based Representations for Facial Emotion Recognition,” Proc. Int. Jt. Conf. Neural Networks, Vol. 2022-July, No. ii, 2022, DOI: 10.1109/IJCNN55064.2022.9891956.

Q. Li et al., “Optimizing Class Imbalance in Facial Expression Recognition using Dynamic Intra-Class Clustering,” Biomimetics, Vol. 10, No. 5, pp. 1–19, 2025, DOI: 10.3390/biomimetics10050296.

N. K. Zuhal, D. P. Pamungkas, and R. Wulaningrum, “Klasifikasi Emosi pada Wajah dengan menggunakan K-MEANS Clustering dan KDEF,” Pros. SEMNAS INOTEK (Seminar Nas. Inov. Teknol., Vol. 5, No. 1, pp. 243–248, 2021.

A. M. Betru, T. M. D. Tran, and W. Ectors, “Automating Composition of Origin-Destination Flows of Intersections based on UAV Data,” Procedia Comput. Sci., Vol. 257, pp. 233–240, 2025, DOI: 10.1016/j.procs.2025.03.032.

C. Shen, L. Qian, and N. Yu, “Adaptive Facial Imagery Clustering via Spectral Clustering and Reinforcement Learning,” Appl. Sci., Vol. 11, No. 17, 2021, DOI: 10.3390/app11178051.

M. Piao Tan and C. A. Floudas, “Determining the Optimal Number of Clusters,” Encycl. Optim., Vol. 1, pp. 687–694, 2008, DOI: 10.1007/978-0-387-74759-0_123.

H. Hendrik, K. Kusrini, and K. Kusnawi, “Optimasi Penentuan Sentroid Awal pada K-Means untuk Meningkatkan Hasil Evaluasi Davies-Bouldin Index,” J. Inform. Teknol. dan Sains, Vol. 6, No. 1, pp. 52–57, 2024, DOI: 10.51401/jinteks.v6i1.3873.

V. I. Agughasi and M. Srinivasiah, “Semi-Supervised Labelling of Chest X-Ray Images using Unsupervised Clustering for Ground-Truth Generation,” Appl. Eng. Technol., Vol. 2, No. 3, pp. 188–202, 2023, DOI: 10.31763/aet.v2i3.1143.

F. A. G. Zhafirah, R. Rokhana, R. Sigit, and B. S. B. Dewantara, “Fusi Algoritma K-Means dan CNN untuk Klasifikasi Emosi pada Anak,” Techno.Com, Vol. 22, No. 3, pp. 622–634, 2023, DOI: 10.33633/tc.v22i3.8667.

D. Evangeline and A. Parkavi, “Facial Emotion Recognition of Online Learners using a Hybrid Deep Learning Model,” Int. J. Intell. Eng. Syst., Vol. 17, No. 6, pp. 735–751, 2024, DOI: 10.22266/ijies2024.1231.56.

L. Manduchi, K. Chin-Cheong, H. Michel, S. Wellmann, and J. E. Vogt, “Deep Conditional Gaussian Mixture Model for Constrained Clustering,” Adv. Neural Inf. Process. Syst., Vol. 14, No. NeurIPS 2021, pp. 11303–11314, 2021.

X. Lin, X. Yang, and Y. Li, “A Deep Clustering Algorithm based on Gaussian Mixture Model,” J. Phys. Conf. Ser., Vol. 1302, No. 3, 2019, DOI: 10.1088/1742-6596/1302/3/032012.

C. Mejia-Escobar, M. Cazorla, and E. Martinez-Martin, “Towards a Better Performance in Facial Expression Recognition: A Data‐Centric Approach,” Comput. Intell. Neurosci., Vol. 2023, No. 1, 2023, DOI: 10.1155/2023/1394882.

H. M. Yisihak and L. Li, “Advanced Face Detection with YOLOv8 : Implementation and Integration into AI Modules,” Vol. 11, 2024, DOI: 10.4236/oalib.1112474.

A. Budi, S. Suma’inna, and H. Maulana, “Pengenalan Citra Wajah sebagai Identifier menggunakan Metode Principal Component Analysis (PCA),” J. Tek. Inform., Vol. 9, No. 2, pp. 166–175, 2018, DOI: 10.15408/jti.v9i2.5608.

F. Nuraeni, H. Susilawati, and Y. Handoko Agustin, “Perbandingan Implementasi Algoritma K-Means++ dan Fuzzy C-Means pada Segmentasi Citra Wajah,” JuTI “Jurnal Teknol. Informasi,” Vol. 1, No. 2, p. 47, 2023, DOI: 10.26798/juti.v1i2.722.

Agyztia Premana, Raden Mohamad Herdian Bhakti, and Dimas Prayogi, “Segmentasi K-Means Clustering pada Citra menggunakan Ekstraksi Fitur Warna dan Tekstur,” J. Ilm. Intech Inf. Technol. J. UMUS, 2020.

K. Dbscan and Y. Hasan, “Pengukuran Silhouette Score dan Davies-Bouldin Index pada Hasil Cluster,” Vol. 06, No. 01, pp. 60–74, 2024.

I. T. Utami, F. Suryaningrum, and D. Ispriyanti, “K-Means Cluster Count Optimization with Silhouette Index Validation and Davies Bouldin Index (Case Study: Coverage of Pregnant Women, Childbirth, and Postpartum Health Services in Indonesia in 2020),” BAREKENG J. Ilmu Mat. dan Terap., Vol. 17, No. 2, pp. 0707–0716, 2023, DOI: 10.30598/barekengvol17iss2pp0707-0716.

Y. Wang, “Improving Spectral Clustering using Spectrum-Preserving Node Aggregation,” Proc. - Int. Conf. Pattern Recognit., Vol. 2022-Augus, No. 30, pp. 3063–3068, 2022, DOI: 10.1109/ICPR56361.2022.9956605.

M. Alshammari and M. Takatsuka, “Approximate Spectral Clustering with Eigenvector Selection and Self-Tuned K,” Pattern Recognit. Lett., Vol. 122, pp. 31–37, 2019, DOI: 10.1016/j.patrec.2019.02.006.




DOI: https://doi.org/10.32520/stmsi.v14i6.5668

Article Metrics

Abstract view : 6 times
PDF - 3 times

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.