Usage Instructions#
For reference examples, please see the Examples section in the lower right corner. Only some parameters are explained here.
1.
When using a public digital human model, avatarSourceFrom=1 and aiAvatarId must be passed.
2.
When using an existing private digital human model, avatarSourceFrom=2 and aiAvatarId must be passed.
3.
When using a locally uploaded video as the digital human template, avatarSourceFrom=1 and videoFileId must be passed.
4.
When using a locally uploaded image as the digital human template, avatarSourceFrom=3, imageFileId, and modeType=1 must be passed.
5.
When using local audio to drive the digital human, audioSourceFrom=0 and audioFileId must be passed.
6.
When using text-to-speech (TTS) to drive the digital human, audioSourceFrom=1 and ttsText must be passed.
6.1 When using a public digital human model, voiceoverId can be omitted to use the voice of the public digital human model;
6.2 In other cases where TTS is used, voiceoverId must be passed.
7.
Other parameters can be optionally passed as needed.
1.
When using video + audio/text without saving the digital human, the parameter captionId is invalid.
2.
When using an image, the digital human will definitely be saved, and the isSave2CustomAiAvatar parameter is invalid.
Supplementary Notes#
2.
The ethnicity ID is obtained from the Ethnicity Query , with the scene parameter set to video_avatar.