A Multimodal Emotion Recognition Framework Based on Generative Adversarial Networks

In recent years, with the rapid development of social media, a large number of open-source emotion recognition datasets have emerged online, providing abundant resources for emotion recognition. Meanwhile, with the help of deep learning, multimodal emotion recognition technology has advanced rapidly. However, deep learning-based multimodal emotion recognition still has some drawbacks, such as requiring a large amount of training data, inconsistencies in the quantity and quality of data within datasets, and heterogeneity among different modalities in multimodal fusion. To address these issues, we propose a multimodal emotion recognition framework based on Generative Adversarial Networks (GANs), which enhances feature representation and reduces heterogeneity in multimodal fusion. Finally, we achieve promising experimental results on the RAVDESS dataset.