Design And Implementaion Of Intelligent Virtual Hunman Generation Engine

In the 21 st century,with the update and iteration of smart phones and GPU hardware,computer graphics has been widely used in life and engineering.Mobile games rely on rendering algorithms to obtain good visual effects,and advanced rendering effects can even simulate the appearance of real people."Intelligent virtual human" is the product of the combination of 3D rendering and artificial intelligence,which is represented as an interactive virtual character that combines intelligent algorithms and has a certain ability of independent learning.The research of virtual humans can provide a reference for exploring human's own intelligence.At the same time,virtual humans have certain application value and can be used for intelligent services,virtual anchors,etc.With the development of 3D-related artificial intelligence technology,some networks can quickly estimate the posture of the character from the video,which provides the possibility for the interaction of intelligent characters and autonomous learning of action resources.For intelligent development of 3D characters,you can choose a large 3D engine,such as Unreal Engine,but there are certain limitations: redundant functions,large footprint,abstract source code,complex architecture,etc.The self-made engine has a small size and flexible customization,which is more conducive to algorithm research.This thesis takes the combination of deep learning and recognition algorithms to control and display virtual characters as the core,designs and implements a lightweight intelligent character development engine,which combines Qt and Imgui interface libraries,and build deferred rendering pipeline with Direct X12.The rendering function can be customized quickly.The engine develops components through the entity component system to meet the scalability and extensibility of different control systems.This thesis designs a set of intelligent control framework,through the recognition of external input,the recognition data is converted into driving data in the driving layer,and finally the driving data is broadcast to various control components to realize intelligent role control.Finally,for gesture simulation and voice control.In the engine,an interactive virtual character is realized.The character can learn actions through external video input and call actions through voice input.This article first introduces the research and application background of some interactive intelligent technologies and 3D engines,expounds the current research status at home and abroad,and gives a summary of the contents of each chapter.Then gives the engine's demand analysis and overall design,and introduces the control framework's design and implementation.Next,we will introduce the related technologies of 3D character animation rendering,introduce and implement the related technologies of each module,including: component module,rendering module,animation module,scene management module,and resource management module.Finally,the function test of each module of the engine is carried out,we summary the work of this thesis,and prospect the follow-up development.