Comprehensive Analysis of Human Emotions using Artificial Intelligence Techniques

Identifier:  TFG:8157

Handle: https://hdl.handle.net/20.500.11797/TFG8157

Authors:  Pérez Gil, Valeria

Abstract:
Emotions are vital to both learning and teaching processes, directly impacting students’ academic performance and teachers’ motivation. This work explores the feasibility of using physiological signals for emotion recognition through artificial intelligence techniques, particularly in smart classroom environments. Specifically, it characterises changes in electrodermal activity, heart rate, and skin temperature during emotional experiences to extract features that describe various emotional states. Experiments using two real-world datasets indicate that relying solely on physiological signals is insufficient for accurately distinguishing emotions, highlighting the need to complement them with additional information, such as facial expressions, posture, and contextual information. Our findings, in addition to contributing to academic research on the use of physiological signals for emotion recognition, could open the door for the development of complex tools for educators, enabling them to adjust their pedagogical methodologies in real-time according to the emotional needs of the students.