JIITA, Vol.9 No.3 pp.1122-1129 (2025), DOI: 10.22664/ISITA.2025.9.3.1122
Vikas Jangra, and Sumeet Gill
Abstract. Emotion is a way humans express their feelings, typically through facial expressions, body language, and tone of voice.
Among these, facial expressions are the most powerful, natural, and universal signals for conveying emotional states.
However, recognizing emotions based on facial expressions can be challenging due to the similarities between certain expressions.
For example, fear and surprise often share overlapping patterns, making it difficult to distinguish them with the naked eye.
This study focuses on developing a mobile-based application for realtime emotion recognition using facial expressions.
It employs a Deep Learning approach, specifically a Convolutional Neural Network (CNN), with the MobileNet algorithm utilized to
train the recognition model. The application is designed to identify four types of emotions: happy, sad, surprised, and disgusted.
The results demonstrate a high level of accuracy, achieving state-of-the-art performance.
Future improvements could involve expanding the range of facial expression categories or exploring alternative deep learning
approaches to enhance the model’s capabilities.
Keywords; Emotion Recognition, Facial Expressions, Convolutional Neural Networks (CNN), MobileNet Algorithm, Real-Time Application
Fullpaper: