Volume 2 - Issue 2, March - April 2026

The significance of emotional recognition in social media videos has grown with the rise of video-based content. This project aims to develop a system that uses computer vision and deep learning algorithms to analyze facial expressions, body language, and auditory cues to identify emotions. The model employs Convolutional Neural Networks (CNNs) to extract visual data and Transformers or Recurrent Neural Networks (RNNs) to capture temporal emotional states. Emotions such as happiness, sadness, anger, and surprise are detected by examining these traits. The system also uses natural language processing (NLP) and speech recognition to analyze audio, improving the accuracy of emotion classification. This multimodal approach, combining video, audio, and facial expressions, provides a comprehensive understanding of emotions, making the system valuable for social media content moderation, user engagement analysis, and personalized recommendations. Additionally, a multi-view feature fusion technique enhances emotion identification by extracting Imaging Photo Plethysmo Graphy (IPPG) signals from facial videos to obtain heart rate variability (HRV). This integration of biometric data with CNN-based visual analysis creates a more resilient model for recognizing emotions. Compared to methods relying solely on facial expressions, this approach significantly improves accuracy. Overall, the system contributes to affective computing by offering a real-time method for emotion recognition, with applications in fields like artificial intelligence, computer science, and psychiatry, where understanding human emotions is critical. By leveraging multimodal data, the system enhances the accuracy and reliability of emotion detection in videos.

Gopika S, Dr.S.Malathi,"Facial Expression Based Emotion Detection in Real Time" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(2): Page(798-801) Mar-Apr 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.