Method and device for generating 3D interactive digital human and customer service project system
By integrating the large language model and the speech model, synchronous matching of speech and lip shape is achieved, the problem of 3D digital human communication in the existing technology is not realistic enough, the realistic and naturalness are improved, and cross-platform interaction is realized.
CN119169161APending Publication Date: 2024-12-20YAOMO
Patent Information
- Application Number
- CN202411233779.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-12-20
AI Technical Summary
Technical Problem
The synchronization relationship between voice and lip shape is not considered in the existing 3D digital human technology, resulting in insufficient communication.
Method used
By integrating large language models and speech models, combining speech recognition, text-to-speech and lip service algorithms, synchronous matching and combination of speech and lip shaped shapes are achieved, target facial forms are generated, and more vivid and realistic 3D interactive digital people are displayed.
Benefits of technology
It improves the realistic and naturalness of 3D digital human communication, enhances the user's visual experience, and realizes cross-platform applications through web links, expanding the scope of interaction.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN119169161A_ABST
Abstract
The invention provides a 3D interactive digital human generation method and device and a customer service project system, and relates to the technical field of computers. The method comprises the following steps: logging in a customer service project system, and creating an initial 3D character model in response to a character role and virtual scene selection operation; according to the virtual scene, responding to editing operation on the initial appearance form and the initial face form, triggering an integrated large language model, and when login succeeds, connecting the virtual scene with the integrated large language model; according to the virtual scene, in response to a selection operation on a voice style of the initial 3D character model, adjusting the initial appearance form according to the voice style to obtain a target appearance form; performing mouth shape matching on the voice style by using the voice model, and combining the matched mouth shape with the initial face shape to obtain a target face shape; project publishing is carried out according to the target appearance form and the target face form, the 3D interactive digital human is displayed, and voice service and mouth shapes can be synchronized into a whole.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Cited By
3D digital human system integration method and system based on credential environment
CN120894472A
3D digital human system integration method and system based on xinchuang environment
CN120894472B