SubmitAvatarVideoJob

• This operation is an asynchronous operation. After a node is committed, a node ID is returned. The node is not completed immediately and enters a background queue to execute asynchronously. The final result is sent through a callback notification. You can also call the Get smart node task result operation to query the node status.

• Currently, only text or human voice audio is supported as input. Audio files support only the MP3 or WAV format.

• Both MP4 and WebM output formats are supported. When the output format is MP4, the final video has a green screen background, and a mask video of the virtual human is also generated for subsequent use. This method is recommended. When the output format is WebM, the final video is a WebM file with a transparent channel, which is active for frontend display. However, the synthesis is slower due to WebM encoding issues.

• The output result includes sentence UNIX timestamps of the human voice for subsequent use in editing.

InputConfig example:

Input as an audio URL:

{
  "InputFile": "https://your-bucket.oss-cn-shanghai.aliyuncs.com/xxx.mp3"
}

Input as text:

{
  "Text": "Do you not see the Yellow River's waters descending from heaven, rushing to the sea never to return.Do you not see in the bright mirror of the high hall, the sad sight of white hair,Hair turns from black to white in the course of a day."
}

Input as a MediaId:

{
  "MediaId": "4aef0c80cc0071edbf92f6e7c44b6302"
}

EditingConfig example:

AvatarId: the ID of the digital human avatar. Required. Valid values: Official digital human avatars.
Voice: the voice actor. Valid only when the input is text. Optional. Valid values: Intelligent speech effect examples.
CustomizedVoice: the customized voice. If you need to use a customized voice generated through voice cloning as the voice actor, use this field instead of the Voice field.
LoopMotion: specifies whether the first and last frames are identical. If this parameter is set to true, the first and last frames of the generated digital human video are the same. When the video is played in a loop, there is no sudden jump in the movements and expressions of the character. Valid values: true and false. Default value: false.
SpeechRate: the speech rate. Valid only when the input is text. Valid values: -500 to 500. Default value: 0.
PitchRate: the pitch. Valid only when the input is text. Valid values: -500 to 500. Default value: 0.
Volume: the volume. Valid only when the input is text. Valid values: 0 to 100. Default value: 50.
BackgroundUrl: the background image. Optional. If this parameter is not specified, a black background is used by default. Only the JPG and PNG formats are supported. The resolution must be 1080p. If a portrait avatar is used, the resolution must be 1080p portrait. If a landscape avatar is used, the resolution must be 1080p landscape.

{
  "AvatarId": "xinxin-marketing_standing", 
  "BackgroundUrl": "https://your-bucket.oss-cn-shanghai.aliyuncs.com/xxx.jpg",
  "Voice" : "zhichu",
  "LoopMotion": true,
  "SpeechRate": 100,
  "PitchRate": 10, 
  "Volume": 10 
}

OutputConfig example:

{
  "MediaURL": "https://your-bucket.oss-cn-shanghai.aliyuncs.com/xxx.mp4"
}

You can call the Get smart node task result operation to query the node execute status and task result. The following example shows the response when the node executes successfully:

{
	"RequestId": "2014D1A8-4143-164F-94B4-32B8F39B706D",
	"JobId": "d9367da8c7184ec7a3f24de530ac5b9a",
	"State": "Finished",
	"SmartJobInfo": {
		"Title": "default_title_2023-03-28T13:37:47Z",
		"EditingConfig": "null",
		"JobType": "AvatarVideo",
		"CreateTime": "2023-03-28T13:37:47Z",
		"ModifiedTime": "2023-03-28T13:37:47Z",
		"UserId": 1833202230108227,
		"outputConfig": {
			"mediaUrl": "https://oushu-test-shanghai.oss-cn-shanghai.aliyuncs.com/avatar/222.mp4"
		}
	},
	"JobResult": {
		"MediaId": "4aef0c80cc0071edbf92f6e7c44b6302",
		"AiResult": "{\"outputVideoUrl\":\"https://oushu-test-shanghai.oss-cn-shanghai.aliyuncs.com/avatar/222.mp4\",\"subtitleClips\":\"[{\\\"from\\\":0.0,\\\"to\\\":4.692,\\\"content\\\":\\\"Do you not see the Yellow River's waters descending from heaven, rushing to the sea never to return.\\\"},{\\\"from\\\":4.692,\\\"to\\\":9.061,\\\"content\\\":\\\"Do you not see in the bright mirror of the high hall, the sad sight of white hair,Hair turns from black to white in the course of a day.\\\"}]\"}"
	}
}

The format of AiResult is as follows:
outputVideoUrl: the URL of the output final video, in MP4 format.
subtitleClips: the sentence timestamps of the subtitles.

{
	"outputVideoUrl": "https://oushu-test-shanghai.oss-cn-shanghai.aliyuncs.com/avatar/222.mp4",
	"subtitleClips": "[{\"from\":0.0,\"to\":4.692,\"content\":\"Do you not see the Yellow River's waters descending from heaven, rushing to the sea never to return.\"},{\"from\":4.692,\"to\":9.061,\"content\":\"Do you not see in the bright mirror of the high hall, the sad sight of white hair,Hair turns from black to white in the course of a day.\"}]"
}