Skip to navigation
Transcriptions API

Cut a face track for a segment

Cuts frames [startFrame, endFrame) at fps from a matched speaker’s face track and returns a bounding-boxes file for that segment. Pass the url as the segment’s active_speaker_detection.bounding_boxes_url so the segment lipsyncs that speaker. url is null when the face is on none of those frames; leave that segment unlipsynced or use auto-detect. 60 requests per minute.

Authentication

x-api-keystring
API Key authentication via header

Path parameters

idstringRequired
Transcription job id.

Request

This endpoint expects an object.
revisionstringRequired
The revision of the transcription job.
speakerIdstringRequired
Id of a matched speaker.
startFrameintegerRequired
First frame of the segment.
endFrameintegerRequired
Frame after the last frame of the segment. Must be greater than startFrame.
fpsdoubleRequired
Frame rate of your source frame grid.
skipobjectOptional
Frames inside the segment that its footage no longer holds, for example because a cut removed them. Must lie strictly inside the segment.

Response

Face track slice created
urlstringOptional

URL of a JSON file with a bounding_boxes array, one entry per frame of the segment. Pass it as the segment’s active_speaker_detection.bounding_boxes_url. Null when the face is on none of the segment’s frames.

Errors

400
Bad Request Error
401
Unauthorized Error
404
Not Found Error
409
Conflict Error
422
Unprocessable Entity Error
429
Speaker Analysis Too Many Requests Error
500
Internal Server Error