Deep Learning-Based Facial Expression Recognition for Analyzing Visitor Engagement

Grace Setiaputri, Jason Chrisbellno Mackenzie, Ivan Sebastian Edbert, Derwin Suhartono

2025

Abstract

Interpersonal communication relates to facial expression because it's a natural source. In the real world, people use their facial expressions, particularly when conducting customer satisfaction surveys for business purposes. Due to the increasing number of fake reviews, evaluating customer satisfaction based on online reviews is sometimes inaccurate. This study uses three machine learning models, ResNet50, VGG16, and EfficientNetB3, to classify human facial expressions. The FER-2013 dataset is used, then oversampled and augmented, the performance of three models was compared using accuracy and F1-Score as comparison values. EfficientNetB3 gets the highest accuracy of 85.41% and F1-score of 85.34%. Future research should apply more sophisticated data balancing techniques, such as the Synthetic Minority Over-sampling Technique (SMOTE), to address data imbalances without adding processing time. Furthermore, extending the number of epochs and refraining from early stopping strategies may help in determining the model's maximal accuracy potential.

Paper Citation

in Harvard Style

Setiaputri G., Chrisbellno Mackenzie J., Sebastian Edbert I. and Suhartono D. (2025). Deep Learning-Based Facial Expression Recognition for Analyzing Visitor Engagement. In Proceedings of the 1st International Conference on Research and Innovations in Information and Engineering Technology - Volume 1: RITECH; ISBN 978-989-758-784-9, SciTePress, pages 32-36. DOI: 10.5220/0014268200004928

in Bibtex Style

@conference{ritech25,
author={Grace Setiaputri and Jason Chrisbellno Mackenzie and Ivan Sebastian Edbert and Derwin Suhartono},
title={Deep Learning-Based Facial Expression Recognition for Analyzing Visitor Engagement},
booktitle={Proceedings of the 1st International Conference on Research and Innovations in Information and Engineering Technology - Volume 1: RITECH},
year={2025},
pages={32-36},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0014268200004928},
isbn={978-989-758-784-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Innovations in Information and Engineering Technology - Volume 1: RITECH
TI - Deep Learning-Based Facial Expression Recognition for Analyzing Visitor Engagement
SN - 978-989-758-784-9
AU - Setiaputri G.
AU - Chrisbellno Mackenzie J.
AU - Sebastian Edbert I.
AU - Suhartono D.
PY - 2025
SP - 32
EP - 36
DO - 10.5220/0014268200004928
PB - SciTePress