Real-Time Facial Emotion Recognition Using a Computationally Efficient CNN and Haar Cascade Face Detection

Facial Emotion Recognition (FER) has emerged as an important research area in artificial intelligence and computer vision due to its applications in healthcare, intelligent tutoring systems, human--computer interaction, surveillance, and driver monitoring. Although deep learning techniques have significantly improved recognition performance, many existing models require high computational resources, limiting their deployment on resource-constrained devices.

This paper presents a computationally efficient Convolutional Neural Network (CNN)-based facial emotion recognition framework integrated with Haar Cascade face detection for real-time emotion classification. The Haar Cascade classifier is employed to detect facial regions efficiently before emotion recognition, reducing computational overhead while maintaining reliable performance. The proposed CNN model classifies facial expressions into seven basic emotions: Angry, Disgust, Fear, Happy, Neutral, Sad, and Surprise using the FER2013 dataset. Data preprocessing and augmentation techniques are applied to improve model robustness and generalization.

The proposed framework aims to balance recognition accuracy, computational efficiency, and inference speed, making it suitable for deployment on standard CPU-based systems and web-based applications. Experimental results demonstrate the effectiveness of the proposed approach for practical real-time facial emotion recognition.