Large Language Model Based Intelligent Interaction for Digital Human

With the rapid development of new technologies such as high-speed Internet, cloud computing, edge computing, etc., the application scenarios of digital human continue to expand, which puts forward higher requirements for the intelligent interaction capability of digital human. This paper presents a study on the large language model (LLM) based intelligent interaction method for digital human and its web applications. Based on the digital human model built in Unity, the intelligent interaction capability is achieved by embedding BaiduAI speech synthesis and LLM ERNIE bot. To verify the proposed intelligent interaction method, an official college website is established, in which the proposed digital human is embedded to act as the intelligent online consultant. From the validation, the proposed digital human could present intelligent and natural answers by both voice and text, effectively realizing the intelligent interaction based on LLM and guaranteeing good experience in web application scenario.