Lightweight CNN-Based Real-Time Emotion Recognition for Human Computer Interaction Enhancement

Venkateswarlu Sunkari, V. Sumathi, Pankaj Naik, P. Usharani, G. Nagarjunarao, Ramabathina Hemanth Kumar

2025

Abstract

Facial expression recognition is essential in achieving natural interactions between human and machine. In this study, we present an efficient CNN framework for real-time facial emotion recognition in human-computer interaction (HCI) systems. In contrast with classical models which are affected by a high latency when dealing with real sequences, ad-hoc CNN architecture has been incorporated, which has low computational load, hence an efficient and accurate emotion classification that can be achieved in dynamic and low-resource conditions. The framework is tested on variety of datasets and real time video sequences to showcase its strength to occlusions, variation in light, and gradual emotional change. Experimental results demonstrate that the system is promising for practical haptic applications in HCI systems, e.g., virtual assistant, smart classroom, or interactive kiosk.

Paper Citation

in Harvard Style

Sunkari V., Sumathi V., Naik P., Usharani P., Nagarjunarao G. and Kumar R. (2025). Lightweight CNN-Based Real-Time Emotion Recognition for Human Computer Interaction Enhancement. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 342-349. DOI: 10.5220/0013863700004919

in Bibtex Style

@conference{icrdicct`2525,
author={Venkateswarlu Sunkari and V. Sumathi and Pankaj Naik and P. Usharani and G. Nagarjunarao and Ramabathina Kumar},
title={Lightweight CNN-Based Real-Time Emotion Recognition for Human Computer Interaction Enhancement},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25},
year={2025},
pages={342-349},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013863700004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25
TI - Lightweight CNN-Based Real-Time Emotion Recognition for Human Computer Interaction Enhancement
SN - 978-989-758-777-1
AU - Sunkari V.
AU - Sumathi V.
AU - Naik P.
AU - Usharani P.
AU - Nagarjunarao G.
AU - Kumar R.
PY - 2025
SP - 342
EP - 349
DO - 10.5220/0013863700004919
PB - SciTePress