Volume 2 - Issue 1, January - February 2026

Emotion recognition is an emerging research area that lies at the intersection of artificial intelligence, psychology, and human–computer interaction. Traditional emotionrecognition systems rely on unimodal inputs such as text, audio, or visual cues, which often leads to limited accuracy and reduced robustness in real-world scenarios. To address these limitations, this paper presents a real-time multimodal emotion recognition system that integrates facial expressions from video input, vocal characteristics from audio signals, and semantic information from textual data to predict human emotions more accurately. The proposed system employs deep learning models tailored to each modality. Convolutional Neural Networks (CNNs) are used for facial emotion recognition, Mel-spectrogram-based representations combined with recurrent neural networks are applied for speech emotion analysis, and word-embedding techniques with LSTM-based architectures are utilized for text emotion classification. A unified web-based application developed using the Flask framework enables real-time input processing and emotion prediction. By combining complementary information from multiple human behavioral cues, the multimodal architecture enhances emotional understanding and improves classification consistency compared to unimodal systems. Experimental observations demonstrate that the system performs reliably across different input modalities, highlighting its suitability for real-time applications such as virtual interviews, educational platforms, mental-health monitoring, and human–computer interaction systems. This work emphasizes the importance of multimodal integration and establishes a strong foundation for future advancements in deep learning–based affective computing.

Dr.S.Seetha, Manoj S V, Pradyumna M P, Samson Anthony,"Real-Time Multimodal Emotion Recognition" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(1): Page(590-598) Jan-Feb 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.