基于AI+动作捕捉技术的虚拟主播体感交互系统的设计与实现

Design and Implementation of Virtual Host Somatosensory Interactive System Based on AI+Motion Capture Technology

随着科技的快速发展和解放生产力的必然要求,人工智能(AI)的应用日益广泛,通过AI技术来推动广播电视领域在内容生产、终端服务和安全监管等方面的融合创新,实现全业务、全流程、全网络从数字化向智能化的战略转型意义重大。本文对"AI+广电"应用领域中的"虚拟主播"进行设计与实现,介绍了人物数字化过程、AI人脸表情跟踪计算方法、交互式虚拟主播3D建模、实验结果及分析。

With the rapid development of science and technology and the inevitable requirement of emancipating productive forces, artificial intelligence has been applied more and more widely. It is of great significance to promote the integration and innovation of content production, terminal services and security supervision in the field of radio and television through AI technology, and to realize the strategic transformation of the whole business, whole process and whole network from digitalization to intellectualization. This paper designs and implements a "virtual host" in the application field of "AI + Radio and Television". It introduces the process of digitizing characters, the calculation method of AI facial expression tracking, interactive virtual host 3 D modeling, experimental results and analysis.