Research On Autonomous Interaction Models Of Virtual Avatar

The users can communicate with quasi face-to-face form in virtual environment. The users can understand intentions of the other side more effectively by virtual avatars. This study based on virtual avatar has become a hotspot and also has broad application prospects. The technology of virtual avatar can be applied to daily real-time communication software. We can communicate with other remote users in order to enhance interaction's naturalness. Because actions of virtual avatars are controlled by the user's actions, namely, the user has operation, the virtual avatar action; whereas, the virtual avatar hasn't action along with the user has no operation which make the virtual avatars lack of autonomous interaction. In virtual environment , if one of the users has no operation, namely, the user fall into his/her operation silences his/her avatar cannot indicate the real intentions and emotions because of the avatar falling into operation silence too. This has resulted in interruption of interaction and cannot maintain continuous,natural of communication which does not meet the state of people's true communication with each other in daily lives.To make the whole process more realistic, in this paper we will analyze and design the virtual avatar for the state of the avatars, and make the avatars autonomous interaction when one user fall into his/her silence. The main work of this paper is as follows:Firstly, this thesis analyzes the interaction models between persons in real life, especially, in the conversation of real life, one user has no response on the other (that is , the user fall into silence ), the user show his/her facial expressions,body movements and so on contain the non-verbal information, as well as the other user's reaction including language information and non-verbal information.Secondly, the algorithms of the silence extent when the user fall into silence are attempted to proposed through the interaction models between persons in real life, and verify the algorithm reasonably by statistic. The user's typical response when he/her falls into silence transplant to the avatar. We design the autonomous interaction of virtual avatar.Thirdly, 3D virtual avatar model including motion control algorithm of body movements and facial expressions are studied. Direct kinematics is made use of to substitute for the transformation matrix of avatar and establish its kinematical equation.Fourthly, by building the communication system based on test–to- visible speeches we can test the autonomous interaction model of the avatar when the user fall into silence. The results showed that: this model maintains the continuity of the system.