Semantic fidelity-oriented virtual conference method and three-dimensional virtual conference system

Abstract

The invention relates to a semantic fidelity-oriented virtual conference method and a three-dimensional virtual conference system. The method comprises the following steps of: forming a tree semantic architecture according to a semantic concept; setting a default signal corresponding to each semantic concept of different hierarchies in the tree semantic architecture; extracting semantic description of a scene where the participants are located from the multi-modal signal and updating a private hierarchical semantic library; updating the public hierarchy semantic library by utilizing the private hierarchy semantic library; the server side sends hierarchical semantic description composed of all the contents in the public hierarchical semantic library to all the user sides, and the corresponding private hierarchical semantic library is updated at each user side according to the hierarchical semantic description composed of all the contents; generating a virtual conference scene in the view angle of the participants; and generating a virtual conference scene in the view angle of the participants in the conference proceeding process. According to the method, the problems that virtual image expressions and actions are rough, a conference scene is not real, and people and the scene are difficult to interact are solved, the virtual conference scale is increased, and the sense of reality of conference participation is improved.

Description

technical field [0001] The invention belongs to the technical field of virtual meeting, and in particular relates to a virtual meeting method and a three-dimensional virtual meeting system oriented to semantic preservation. Background technique [0002] Affected by economic globalization and the COVID-19 pandemic, remote conferencing technology has been widely used in government, schools, companies and other scenarios, and it is gradually replacing face-to-face conversations as the mainstream medium for meetings. Currently commonly used teleconferencing systems mainly include teleconferencing systems, video conferencing systems, and multimedia platform sharing systems. However, compared with face-to-face conversations, the existing teleconferencing systems have the problems of lack of realism, immersion, and lack of means of communication for participants. With the advancement of augmented display and virtual reality technology, virtual conference technology has emerged as ...

AI Technical Summary

Problems solved by technology

There are two problems with this solution: 1) In order to make the human visual system feel real, the virtual reality video needs to reach 21K resolution 60 frames per second

At present, the theoretical bandwidth of 5G is only 10Gbps, which cannot meet the needs of normal meetings.

2) The existing algorithms for generating 3D models based on high-definition images are time-consuming and labor-intensive, and existing servers cannot meet the real-time computing power requirements of large-scale conferences

However, the preset virtual image cannot reflect important features such as the appearance and demeanor of the participants, resulting in a lack of realism in the meeting, and even causing resentment among the participants due to the uncanny valley effect

[0005] Therefore, there are two problems in the first solution: 1) In order to make the human visual system feel real, the virtual reality video needs to reach 21K resolution 60 frames per second

At present, the theoretical bandwidth of 5G is only 10Gbps, which cannot meet the needs of normal meetings.

2) The existing algorithms for generating 3D models based on high-definition images are time-consuming and labor-intensive, and existing servers cannot meet the real-time computing power requirements of large-scale conferences

In the second solution, the preset virtual image cannot reflect important features such as the appearance and demeanor of the participants, resulting in a lack of realism in the meeting, and even the problem of disgust among the participants due to the uncanny valley effect

[0007] Therefore, the shortcomings of existing solutions all stem from the failure to achieve semantic fidelity, that is, the failure to achieve semantic characterization, signal semantic extraction, transmission and reproduction, and user intention detection feedback.

Images