Facial Emotion Recognition using Action Units-based Machine Learning

Valdez, Manuel Amir Freer, Ghergulescu, Ioana and Moldovan, Arghir-Nicolae (2026) Facial Emotion Recognition using Action Units-based Machine Learning. In: 4th Cognitive Models and Artificial Intelligence Conference, AICCONF 2026 - Proceedings. IEEE, Prague, Czech Republic, pp. 1-8. ISBN 979-833159203-5

Full text not available from this repository.

Abstract

Facial emotion recognition (FER) has been extensively studied through the lens of deep neural networks, yet the predictive capacity of anatomically grounded features, specifically Action Units (AUs) defined within the Facial Action Coding System (FACS), remains underexplored in the context of classical machine learning. This paper investigates the performance of traditional ML algorithms trained on AU features extracted with the OpenFace tool across four benchmark datasets: CK+, FER+ / FER2013, SFEW and ExpW. Two experiments are conducted. The first experiment evaluates four classifiers — Random Forest (RF), K-Nearest Neighbours (KNN), Naïve Bayes and Support Vector Machine (SVM), trained on AU features. The second experiment evaluates the influence of 255 OpenFace feature set combinations on the classification performance using RF. The results indicate that Random Forest with a combination AU and other facial features achieves 95.1% accuracy on the CK+ dataset. Moreover, the proposed approach outperformed directly comparable AUs + CML based methods from past works on the controlled benchmarks CK+ and SFEW, whilst the lower performance on unconstrained datasets reflects the intrinsic difficulty of in-the-wild FER and the need for more complex DNN-based methods.

Item Type:	Book Section
Uncontrolled Keywords:	Action Units (AUs); Facial Emotion Recognition (FER); Image Classification; Machine Learning; OpenFace
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence > Computer vision Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence > Computer vision B Philosophy. Psychology. Religion > Psychology > Emotions Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Staff Research and Publications
Depositing User:	Tamara Malone
Date Deposited:	12 Aug 2026 13:26
Last Modified:	12 Aug 2026 13:26
URI:	https://norma.ncirl.ie/id/eprint/9521

Actions (login required)

View Item