This article focuses on the practical application and optimal design of human-computer interaction (HCI) technology with edge computing in the intelligent exhibition of museum cultural heritage, and then proposes an intelligent exhibition system architecture based on edge node deployment and resource dynamic scheduling. This architecture reduces the interaction delay and improves the efficiency of multimodal perception and feedback by putting the data processing and artificial intelligence (AI) model reasoning links on the edge devices close to the user. In the research, taking the digital protection project of Dunhuang Mogao Grottoes as an example, 3D laser scanning, deep learning image restoration and VR/AR technology are used to complete the digital collection and virtual reconstruction of cultural relics, and AI models such as ResNet-152 and DCGAN are deployed on edge devices for localization reasoning. Experiments show that compared with the traditional cloud architecture, the average response time of the edge enhancement system is reduced by 68.6%, the user's stay time is increased to 42.6 minutes, and the score of knowledge understanding is 8.5 (out of 10), which significantly improves the audience's participation and cultural cognitive effect. This article puts forward a closed-loop interactive paradigm of "perception-cognition-feedback", which provides theoretical support and practical path for building an intelligent and evolvable wisdom museum.
Hanbing et al. (Thu,) studied this question.