AI digital assistant

Solution Introduction

Based on cutting-edge Faceunity technologies that include virtual image and mouth shape generation, multi-terminal rendering, dual-mode drive and deep learning technology, coupled with the help of APP, smart screen, PC, mini programs, H5, smart speaker and other terminal display forms, we provide users with the quality interaction, consultation, explanation, customer services, guided tours and other services that a real person can offer.

Solution Architecture

Applications

Financial Customer Service

Government Assistant

In-vehicle Assistant

Smart Tour Guide

Medical Assistant

Solution Formats

ios

Android

Mini program

Windows

Linux

Service Methods

Standard SDK

Cloud deployment push

Core Functions

Image creation and rendering

Image selection / customization
3D image rendering in real time
High performance physical simulation

Image drive animation

Mouth and lip drive animation
Dynamic expressions
Responsive movement

Scene interaction

Text broadcasting
TTS
Multi-modal interaction

Underlying technology

Voice Recognition

Semantic Understanding

Text-to-Speech

3D Reconstruction

Physical simulation

Rendering Engine

Intelligent drive

Solution Features

Compatible with mainstream voices

Our avatars can adapt seamlessly to a customer's original TTS, ASR and NLP, and support the conversion of text-to-speech conversations into question-and-answer mode for virtual images.

Image style customization

We allow users to create tailor-made AI avatars, including 3D cartoon people, 3D animals, 3D simulated people, 2.5D virtual avatars, real-life deep learning, etc., so as to meet the needs of customization in a variety of situations.

Voice and mouth shape synchronization

We allow real-time mouth shape synchronization based on the content of Q&A to achieve consistent voice-driven mouth shape synchronization and drive multi-modal interaction with avatars.

Rich expressions and movements

We use unique BlendShape computer graphic synthesis technology to drive and blend the face of the avatar and display realistic expressions. The models can support any movement, which enhances the expressiveness of the image.

Adaptation for Multiple Clients

We allow both online and local rendering, providing a wide range of professional virtual avatars compatible with mini programs, H5 pages, APP, smart screens, in-vehicle assistant, mobile and windows platforms.