Home
Virtual Human
Virtual Content
Core Technology
Business Scenario
Product Center
Three-dimensional virtual content production intelligent cloud platform Virtual live broadcast and offline real-time interactive products AI virtual human capability platform
Three-dimensional virtual content production intelligent cloud platform
Virtual live broadcast and offline real-time interactive products
AI virtual human capability platform
News Center
About us
CN
CN

AI virtual human capability platform

The creation of everyone's AI clone is one of the core AI capabilities of the meta-universe era. The AI virtual human capability platform is based on the industrialized platform for collaborative production of virtual content independently developed by Mofa, and through Mofa's self-developed AI virtual human core technology, including TTSA technology (text-driven speech and animation technology), STA technology ( Voice-driven animation technology), ETTS (stylized speech synthesis technology), intelligent action and expression synthesis technology, etc., combined with MOFA's self-developed or third-party intelligent dialogue system and third-party engine, to achieve a one-stop place for creating, educating, and employing people The ability platform for AI virtual humans is constructed in a manner that empowers developers and operators to create multi-modal interactive AI virtual humans on the platform, which can be applied to different business scenarios.

Virtual customer service

Virtual teacher

Virtual anchor

Virtual doctor

Virtual government

Android App iOS App Windows应用 智慧大屏 小程序 网页应用 ... 软硬一体机

企业

个人开发者

ISV

...

形 象 定 制
技 能 定 制
资 产 定 制
语 音 能 力
理 解 能 力
说 话 能 力
表 情 能 力
动 作 能 力
原生技能库
第三方技能库
自定义技能库

For general users, such as products and operators, the platform provides the ability to build AI virtual human products with zero code, that is, without programming, you can create exclusive AI virtual human products on the platform by creating, educating, and employing people, such as Android App, iOS APP, web application, applet, etc., can be directly applied to various business scenarios without secondary development.


For developers, the platform provides development tools that support multi-platform AI virtual human capabilities applications, such as Android SDK, iOS, SDK, JS SDK, applet SDK, API interface, etc. After developers build virtual humans on the platform, they can use development tools Integrate the capabilities of the AI virtual human into its own products, thereby upgrading the product to support multi-modal interaction AI virtual human products, and apply them in various business scenarios.


One-stop construction of AI virtual human products

The platform provides one-stop ability to build AI virtual human products. For general users, it provides a way to build AI virtual human products with zero code, that is, users can create AI virtual human products without programming development and directly apply them on different platforms; for developers, provide multi-platform AI virtual human development tools, Developers flexibly integrate virtual human capabilities into their products to empower different business scenarios.


Intelligent, efficient and large-scale high-quality avatar customization

Based on the self-developed 3D virtual content production intelligent cloud platform of Mofa, it supports the customization of various styles of high-quality 3D virtual images through intelligent methods, high efficiency, low cost, and large-scale customization, including super-realistic virtual characters and three-dimensional beauty. Characters, cartoon characters, two-dimensional characters, etc.


High-quality real-time intelligent synthesis of voice animation

Based on Morfa’s self-developed text real-time synthesis of speech animation (Text To Speech and Animation), speech synthesis animation (Speech To Animation), stylized speech synthesis technology (Expressive Text To Speech), supports the intelligent synthesis of AI virtual human voice, mouth animation, facial expressions, body movements, etc. through voice or text, so that AI virtual humans can be as vivid and lifelike as real people, and their body movements are natural and smooth.


Multi-modal real person takes over the drive

Based on the self-developed speech synthesis animation (Speech To Animation), a single-camera virtual live broadcast technology, supports real-time acquisition of real people's voice and images, and drives virtual people for multi-modal output. Compared with the intelligent interaction of AI virtual humans, real human takeover allows virtual humans to make more natural and accurate expressions. By combining AI-driven and human-driven methods, virtual humans can be applied to more scenarios.


Multi-system terminal support

AI virtual humans built through the platform can be flexibly applied to different system terminals, such as Android, iOS, Windows, Linux


Flexible deployment method

Support multiple service deployment methods, such as cloud deployment, localized deployment, etc.

应用描述

Based on self-built 3D scanning booth and intelligent modeling and binding technologies, it can quickly and high-quality create virtual images of enterprises, such as virtual spokespersons, virtual trainers, virtual employees, etc. The role images include ultra-realistic, three-dimensional beauty, and two Dimensional and cartoon image. Combining the question-and-answer system with enterprise-related information, create various types of enterprise AI virtual persons that can automatically collect enterprise information, answer questions, and explain enterprise-related knowledge, and improve the effect and efficiency of the enterprise in brand display, brand promotion, and corporate culture development.


Advantage

Super-realistic, beautiful, cartoon, two-dimensional and other types of virtual image design and creation capabilities to create a corporate brand exclusive image;
restore the real population one-to-one, multilingual AI synthesis technology support;
Massive body movement library, automatically synthesizes smooth and natural body movements based on semantics;
supports real-person text, voice, and single camera takeover, and the natural synthesis of body movements during the takeover process can create high-quality, high-precision AI virtual human content. Such as long videos, short videos, pictures, live broadcasts, etc., supporting enterprise applications in multiple occasions


Applicable scene

Corporate training, corporate event hosting, corporate endorsements, corporate brand promotion

·Typical Case

The virtual image of the enterprise serves as the host of the press conference

Application Description

Based on self-developed super-realistic 3D model construction, TTSA (text-driven speech animation synthesis) technology, the existing AI text and voice customer service are upgraded to Multi-modal virtual customer service in the form of video output, restores the real one-to-one video customer service mode, and enhances the service experience of AI customer service.



Advantages

Professional 3D art team, high-precision super-realistic virtual characters to create art pipeline; One-to-one restoration of the real population, multilingual AI synthesis technology support; Massive body movement library, which automatically synthesizes smooth and natural body movements based on semantics; Support real text, voice, single camera takeover, and natural synthesis of body movements during takeover; IOS, Android, HTML5, WeChat applet omni-channel adaptation;



Applicable Scenarios

APP, VoLTE, WeChat applet, smart big screen


·Typical Case

Huaxia Bank Virtual Customer Service

application description

Provide 7*24 hours AI virtual anchor program for brands and e-commerce platforms. Merchants only need to pre-make product introduction text and FAQ corpus. Drive virtual anchors to introduce corresponding products 7*24 hours, and answer user questions during product introductions, helping businesses increase the live broadcast duration and increase the overall conversion rate of the store;


Advantages

High-precision virtual image driving ability, closer to the effect of real people carrying goods, improving stay time and conversion rate; super-realistic, beautiful, cartoon, The ability to design and create multiple types of virtual images such as the second dimension to create a brand-specific image; support multiple driving methods such as AI text drive, live voice drive, and single camera live broadcast drive to enrich the live broadcast form of virtual anchors; create high-precision virtual anchor content , Such as pictures, short videos, etc., to help virtual anchors complete the conversion from tool people to brand IP;


Applicable scenarios

e-commerce platform, private domain traffic platform


·Typical Case

Martin virtual e-commerce anchor

Application Description

Based on its own 3D art R&D team and AI capabilities, it provides enterprises with a one-stop AI virtual teacher from original paintings, 3D modeling to AI interaction Create services and apply them to scenes such as 3D course content production, AI interactive courses, dual-teacher teaching, etc., to help enterprises and teaching institutions improve the efficiency of 3D course content production, get rid of the dependence of enterprises on the image of real teachers, and expand the virtual IP application scenarios in the education industry; /span>


Advantages

Film-level 3D content full-process building team, based on 3D virtual content collaborative production intelligent cloud platform empowered high quality, high efficiency, high productivity 3D asset and content production pipeline; The leading AI-driven algorithm makes the AI virtual teacher's expression more fluent, smarter, and more anthropomorphic; It is equipped with multiple driving methods such as AI text drive, human voice drive, single camera live broadcast drive, etc., with a wider range of application scenarios;


Applicable Scenarios

online education, school education, corporate training


·Typical Case

Bytedance Guagualong Virtual Speaking Teacher

Application Description

Based on the self-built 3D scanning booth and intelligent modeling technology, the 3D virtual avatar of real doctors can be constructed quickly and with high quality, combined with a structured medical knowledge question and answer system, to create an AI virtual doctor that can automatically collect patient information, answer patient questions, and explain medical knowledge Enhance the efficiency of patients' access to medical knowledge and reduce the burden on doctors;



Advantages

Self-built high-precision 3D scanning booth, one-to-one scanning to restore the image of the doctor; AI virtual doctor's mouth shape, body movements, and facial expressions can be deeply customized, restoring the daily expression habits of real doctors; Customized construction of a virtual doctor knowledge system, and continuous iteration ; One-stop interactive video construction platform to quickly build a cross-platform, interactive AI virtual doctor application;



Applicable Scenarios

Medical guidance, information collection, psychotherapy, video visit


·Typical Case

Virtual Therapist at Shanghai Mental Health Center