Research on multimodal human-computer interaction technology based on audiovisual fusion

With the improvement of information technology, service robots are becoming more and more deeply involved in our work and life, and provide a wider variety of services. How to make robots communicate with humans more intelligently and understand and accomplish people's task needs has become a current research hotspot. In this paper, we propose a multimodal human-robot interaction technique with D-S evidence theory for fusion of audiovisual information, and a multimodal interaction system with speech and gesture interaction to improve the performance of human-robot interaction.