• Rushal Chauhan
• Rajeev Mathur

Multimodal Emotion Recognition (MER) is one of the newest merger technologies to date and can be used in affective computing, human-computer interaction, and sentiment analysis. This review examines state-of-the-art deep learning-based approaches to fusion in MER systems, focusing on early and late fusion strategies. This paper attends to the capture of inter-module interactions in text-audio-video recognition systems and its impact on emotion recognition. The need for explainable models, generalization beyond the original corpus, and additional data or modality integration are some of the critical gaps addressed. The review outlines unresolved issues, focusing on few-shot learning and large language models as a means of increasing flexibility and generalization.

Multimodal Emotion Recognition (MER), Early Fusion, Late Fusion, Hybrid Fusion

"Multimodal Emotion Recognition: A Comprehensive Review", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 12, page no.207-215, December-2025, Available :http://www.jetir.org/papers/JETIRHE06026.pdf

"Multimodal Emotion Recognition: A Comprehensive Review", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 12, page no. pp207-215, December-2025, Available at : http://www.jetir.org/papers/JETIRHE06026.pdf