EMOTION RECOGNITION THROUGH FUSION OF HETEROGENEOUS DATA

Computer Engineering

This paper explores the advancement in emotion recognition by fusing heterogeneous data sources, crucial for human-computer interaction and various applications including affective computing and mental health monitoring. Traditional methods relying on single data modalities often lack robustness. Conversely, leveraging diverse data such as facial expressions, speech, physiological signals, text, and contextual information offers promise. The review delves into challenges and opportunities in integrating these modalities and surveys fusion techniques like feature-level, decision-level, and hybrid approaches, assessing their efficacy in capturing complex human emotions. Additionally, it discusses emerging trends like multimodal deep learning and adaptive fusion strategies. Specifically, it explores the fusion of audio and text data, highlighting the complementary information they provide for emotion recognition. Envisioning more context-aware emotion recognition systems with real-world applications in healthcare, education, and human-computer interaction, this paper sets the stage for innovative advancements in the field.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.