Comprehensive Literature Review on Adaptive Multimodal Emotion Recognition Using Deep Learning and Attention-Based Fusion Techniques

Emotion recognition has surely become an important field in artificial intelligence. Moreover, it helps make better communication between humans and computers through emotional computing. As per research findings, human emotions show through face expressions, voice, and written text, so single-method systems are not enough for correct recognition. Regarding emotion detection, multiple ways are needed for better accuracy. Further, as per the progress in deep learning, multimodal emotion recognition is getting much attention regarding its ability to combine different data sources. This study reviews the development of emotion recognition methods as per traditional approaches, machine learning, deep learning, and multimodal systems. The review covers different techniques regarding how emotions can be identified and recognized. The system actually focuses on attention-based fusion methods and adaptive learning that definitely improve performance. Basically, the review shows the same big problems like mixed data