The first look: A biometric analysis of emotion recognition using key facial features

Ana M.S. Gonzalez-Acosta ¹ Marciano Vargas-Treviño ^1* Patricia Batres-Mendoza ¹ Erick I. Guerra-Hernandez ¹ Jaime Gutierrez-Gutierrez ¹ Jose L. Cano-Perez ¹ Manuel A. Solis-Arrazola ² Horacio Rostro-Gonzalez ^2,3*

• ¹ Benito Juárez Autonomous University of Oaxaca, Oaxaca, Oaxaca, Mexico
• ² Department of Electronic Engineering, University of Guanajuato, Guanajuato, Guanajuato, Mexico
• ³ Ramon Llull University, Barcelona, Catalonia, Spain

Facial expressions are pivotal in human emotion recognition and social interaction. While prior studies have underscored the significance of the eyes and mouth, few have validated these findings with robust biometric evidence. This research explores the prioritization of facial features during emotion recognition, introducing a streamlined approach to landmark-based analysis with enhanced efficiency. An experiment was conducted with 30 participants, who evaluated images representing emotions such as anger, disgust, fear, neutrality, sadness, and happiness. Eye-tracking technology was employed to capture participants' gaze patterns, revealing the regions of focus during the recognition process. The findings confirmed a consistent prioritization of the eyes and mouth, with limited attention directed to other facial areas. Based on these insights, we propose a reduced facial landmark model that focuses on the most critical regions for emotion detection, significantly optimizing the traditional 68-point model to only 24 points without compromising accuracy enabling faster processing. This model has been validated using multiple classifiers, including Multi-Layer Perceptron (MLP), Random Decision Forest (RDF), and Support Vector Machine (SVM), demonstrating its robustness across different machine learning approaches.