**Virtual Digital Human Production Technology Core and Challenges in the Virtual Digital Human Manufacturing Process**
The production process of virtual human covers many technologies, and the production method is still gradually developing. The tendency for certain steps to converge with each other is the same, but the complete solution has not yet been fully finalized. Three core and challenging processes in the virtual human manufacturing process: modeling, driver and rendering.
The three modeling costs and technical path differences of the virtual digital human are the costs of the three architectural models of the virtual digital user. At present, the main modeling of virtual digital human has produced three kinds. Depending on the level of human involvement, it is purely manual modeling, building models with the help of collection equipment, and building using artificial intelligence; both related software and hardware are involved, including modeling software, driver software, rendering engines, shooting acquisitions equipment, optical equipment, display equipment, etc. 1. Pure manual modeling method – high cost and low production capacity; 2. Modeling using collection equipment. Moderate cost and broad application; 3. Leverage AI to build models – lower cost building models and techniques need improvement.
The model determines the basic operation of the virtual human in the form of real-life motion capture, that is, the driver renders the model and its motion to display the best effect according to the final effect and application scene of the virtual human determined during the design process and other external factors.
At present, Shiyou Technology combines self-developed motion capture technology, precision algorithm technology, real-time animation technology and AI virtual live broadcast technology. Coupled with cloud real-time rendering, it will present virtual live character images, actions and emojis with high quality, while realizing real-time live broadcast interaction in video and offline scenarios. It has been applied in major scenarios such as broadcast television, education, culture, and entertainment.
In terms of business model, the main service content of Shiyou Technology is software and hardware systems and technical production services, including virtual content customization, original painting design, 3D modeling, rigging, animation production, dynamic capture real-time drivers, real-time lighting materials/rendering = Solving = special effects, post compositing, motion capture equipment rental and sales, technical live execution, virtual studio venue rental.
According to Pan and Lin, creating virtual digital employees mainly requires 3D modeling based on banking, voice library and artificial intelligence technology to build virtual numbers. Because the current technology industry is very developed, the technical threshold for virtual digital employees is not high. Some platforms already provide programs for virtual digital people to develop collaborative platforms. But the real question for virtual digital workers is how to improve the user experience. Early AI has poor experience in responding and improving chatbot responsiveness requires a long-term exploration process.
Virtual digital employees and other functional virtual figures are not the same as virtual idols. The role of the person who creates the idol’s virtual figure is closer to the two-dimensional animation, and its core lies in its IP value and derivative ability. However, the core of functional virtual digital human lies in technical ability. A high level of skill is required for modeling, driving, rendering and artificial intelligence. Creating virtual digital humans also requires high costs, and banks will also face challenges in establishing and using virtual digital employees.
https://www.sohu.com/a/536230955_120624328