The broadcast digital human is suitable for various content production scenarios, including training and media, and serves industries such as media, asset management, and education by supporting custom broadcast content. Using digital human broadcasting enhances human efficiency, reduces costs, and offers low migration and replication expenses. It is minimally affected by human emotions and natural conditions, enabling error-free broadcasting.
Overview of platform capabilities for the broadcast digital human module:
Supports 2D live-action video production with options to edit text content, anchor settings, and video settings. Completed video and audio files can be downloaded. The 2D live-action video production distinguishes between two avatar types: 2D instant avatar and 2D studio avatar.
Supports 3D digital human video production, with options to edit text content, anchor settings, and video settings. Completed video and audio files can be downloaded.
Access path for audio/video broadcasting module: Go to Homepage > Scene Application > Audio and Video Broadcasting to manage your produced audio and video content or create audio/video broadcast projects.
I. Audio/Video Creation
Click the first blank + card, select the avatar type for the audio/video broadcast you want to create, and click Create New Audio and Video to start editing and producing new content.
Select the type of audio/video you want to create.
II. Audio/Video Management
For created audio/video content, hover over the corresponding cover with your mouse pointer to manage the content.
Notes:
The content created with the root account is not visible to sub-accounts, nor edited or deteletd with sub-accounts.
The content created with sub-accounts is viewable to the root account but cannot be edited or deleted with the root account.
Video content includes the following operations:
Re-editing: Allows modifications to edited content. Re-editing does not alter the original video and a new version will be generated.
Video download: Supports downloading in MP4 format and WEBM format (with green-screen avatars only, allowing output with a transparent channel).
Subtitle download: Supports downloading in SRT format.
Video deletion: Removes the video from the platform, and it will no longer be stored.
Headline modification: Renames the video headline.
Headline copy: Copies the video headline.
III. Audio/Video Production
There are three driving capabilities: text-driven, original-voice-driven, and voice-changing-driven. The audio/video broadcast module supports text-driven and original voice-driven methods to produce digital human audio and video files.
Driving Mode
Capability Description
Text-driven
Generates digital human audio and video content with automatically matched mouth shapes by simply inputting text. By inserting action/expression tags within the text, the digital human can perform corresponding expressions and actions at specified points.
Original-voice-driven
Generates digital human audio and video content with automatically matched lip movement by simply inputting audio. The digital human’s voice will match the input audio exactly.
Voice-changing-driven
Generates digital human audio and video content with automatically matched lip movement by inputting audio. The digital human’s voice will match the voice selected during the avatar settings stage.
3.1. Text-Driven Mode
To use the text-driven mode, first select the digital human’s avatar, style, voice, and output settings. Then, enter the text, insert action/expression tags as needed, and check pronunciations for polyphonic characters. This setup will generate a digital human broadcasting video that includes synthesized voice based on the text you provided.
Once the production is completed, click Generate Video, edit the video content name, and select the video format to start the generation process. This process typically takes 1–10 minutes, depending on the length of your video and the broadcasting concurrency purchased for your account. Once the content cover no longer displays a waiting prompt and shows the content normally, you can click to download.
3.2. Audio-Driven Mode
With the audio-driven mode, the generated video will use the uploaded audio file directly without requiring a digital human voice selection. Choose the audio-driven mode to upload an audio file to drive the digital human. Supported five formats include WAV, MP3, WMA, M4A, and AAC.
The remaining digital human style configurations and output settings are the same as those for the text-driven option.
Yes
No
Was this page helpful?