Virtual human assistant solution - Baidu AI open platform

Cooperation consulting

Experience AI

Open Baidu APP

"Sweep" experience immediately

Scheme introduction

Scheme architecture

Application scenarios

Characteristic advantages

Related recommendations

Scheme introduction

Templated Virtual Assistant

Use the virtual assistant template provided by the platform to drive the virtual assistant face according to the input text or voice, so that the virtual assistant has specific expressions and lip movements. At the same time, it can support changing faces for the template. Users can upload the face video of the target person, change faces for the preset template, and generate personalized virtual assistants.

Customized virtual assistant

Upload a 300 second video of the target character, quickly model the visual image of the character, and generate a virtual assistant image. The face of the virtual assistant is driven according to the input text or voice, so that the virtual assistant has specific expressions, lips or body movements.

Scheme architecture

Scheme introduction and product composition

Based on Baidu's leading AI and deep learning technology, it provides a solution to customize the image of virtual visual assistant

Production process

Rapid expansion plan: 1. Upload 300 second face up video of users. 2. Select production configuration: face changing or face changing+body. Fine customization scheme: customized models collect voice+face video data for 2 hours

Usage

Input: text+TTS voice or live voice, output: generated virtual visual assistant video stream

Application scenarios

Interesting interactive video

Virtual anchor broadcast

Interesting interactive video

In the video industry, the ability of virtual assistant can be applied to flexibly generate various interesting videos; It can also integrate the virtual assistant generation capability into the video APP to enrich the video playing methods and attract more users

Cooperation cases

Characteristic advantages

Complete scheme

It provides a complete solution from uploading character videos, changing faces and bodies, to generating customized expressions and synthesizing virtual assistant videos.

Extremely low cost

Users only need to provide 300 seconds of video to complete low-cost modeling in 90 minutes, and generate personalized and configurable virtual hosts and virtual visual assistants

Highly customized

The face, body and synthetic voice all support customization, and the emotion and lip shape of the virtual character can be automatically adjusted according to the input voice information.

Get AI solutions now

Get in touch with us quickly, please click the cooperation consultation below

Cooperation consulting