Genre-Sensitive Prediction of Emotional Arousal in Virtual Reality: A Neural Modeling Approach Using Skin Conductance Peaks
Journal
IEEE Latin America Transactions
ISSN
1548-0992
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Date Issued
2025-12
Author(s)
José Varela-Aldás
Demián Velasco Gómez Llanos
Santiago Arreola Munguía
Marco Antonio Manjarrez Fernandez
Juan Pablo Villaseñor Navares
Type
text::journal::journal article
Abstract
Understanding how different virtual reality (VR) game genres modulate physiological arousal is crucial for designing emotionally adaptive immersive systems. This study introduces a novel experimental framework combining high-resolution Skin Conductance Response (SCR) data and neural predictive modeling to compare emotional activation across horror, skill-based, and exercise VR games. Using Galvanic Skin Response (GSR) sensors, we recorded phasic peaks in SCR from 25 university-aged participants during gameplay sessions with controlled exposure times and standardized transitions. However, given the minimal difference relative to the large variability, this observation should be considered preliminary and specific to the tested games and cohort. A feed-forward neural network was developed to forecast individual arousal levels based solely on genre-induced features, achieving strong predictive performance. This dual contribution empirical genre comparison and lightweight predictive modeling offers a scalable tool for integrating emotional responsiveness into VR systems without continuous biosignal monitoring. The findings not only advance the state of the art in affective computing but also open new avenues for therapeutic, educational, and entertainment applications grounded in physiological adaptation.
License
Acceso Abierto.
URL License
How to cite
C. Del-Valle-Soto et al., "Genre-Sensitive Prediction of Emotional Arousal in Virtual Reality: A Neural Modeling Approach Using Skin Conductance Peaks," in IEEE Latin America Transactions, vol. 23, no. 12, pp. 1356-1364, Dec. 2025, doi: 10.1109/TLA.2025.11231219.
Table of contents
I. Introduction -- II. Related work -- III. Methodology -- IV. Results and discussion -- V. Conclusion.