Author(s)

Bin Gan, Mingyao Zhang, Chaolin Li, Kexu Wu

Corresponding Author

Mingyao Zhang

Abstract

This paper explores the application and optimization of deep learning algorithms that integrate multimodal data in the field of human-computer interaction (HCI) emotion recognition. With the continuous development of artificial intelligence technology, multimodal data processing has become crucial in enhancing the level of intelligence in human-computer interaction. The paper first introduces the fundamental principles of multimodal data fusion, encompassing the integration and comprehensive analysis of various data types such as images, text, and audio. It then elaborates on the application of deep learning algorithms in multimodal emotion recognition, leveraging advanced techniques such as Recurrent Neural Networks (RNNs), Convolutional Neural Networks (CNNs), and self-attention mechanisms to achieve accurate recognition of complex emotional states. Furthermore, the paper analyzes the current challenges faced by multimodal emotion recognition technologies, including the complexity of data fusion, the subjectivity and diversity of emotions, and proposes corresponding optimization strategies, including algorithm and model optimization, personalized emotion model construction, privacy protection mechanisms, and interdisciplinary research. Finally, the paper outlooks the future development trends of deep learning algorithms integrating multimodal data in the field of human-computer interaction emotion recognition, emphasizing their tremendous potential in enhancing user experience and boosting the intelligence of human-computer interactions.

Keywords

Multimodal Data Fusion; Deep Learning Algorithms; Human-Computer Interaction; Emotion Recognition; Recurrent Neural Networks (RNNs); Convolutional Neural Networks (CNNs)