Transcription video

Transcription API converts visual and autio context of the video into text representations. You could transcribe an uploaded vidoe in two ways:

• AUDIO: Transcribing the video's audio content into text.
• VIDEO: Transcribing the video's visual content into text.

Host URL

• https://mavi-backend.memories.ai

Submit Transcription Task

You can submit a transcription task through this interface with the following options:

• Choose between AUDIO or VIDEO transcription.
• Specify a callback address to receive the transcription results automatically.
• Opt not to use a callback—in this case, you can retrieve the transcription results using the query interface.

POST /api/serve/video/subTranscription

Request Example

import requests
headers = {"Authorization": "Bearer <YOUR_ACCESS_TOKEN>"} # access token
data = {"videoNo": "<VIDEO_NO>","type":"AUDIO/VIDEO","callBackURI":"<CALLBACK>"}
response = requests.post(
    "https://mavi-backend.memories.ai/api/serve/video/subTranscription",
    headers=headers,
    json=data
)
print(response.json())

Replace the placeholders in the code above with your actual values:

• YOUR_ACCESS_TOKEN: Your access token.
• VIDEO_NO: The unique video number.
• AUDIO or VIDEO: The transcription type (choose one).
• CALLBACK: (Optional) A publicly accessible URL endpoint,

Request Body

{
  "videoNo": "string",
  "type": "string",
  "callBackURI": "string"
}

Name	Location	Type	Required	Description
Authorization	header	string	Yes	authorization token
videoNo	body	string	Yes	video number
type	body	string	Yes	transcription type: AUDIO, VIDEO
callBackURI	body	string	No	callback address

Response Example

Status code 200

{
  "code": "string",
  "msg": "string",
  "data": {
    "taskNo": "string"
  }
}

Name	Type	Required	Description
code	string	true	response code
msg	string	true	message with response code
data	object	true	JSON data
» taskNo	string	true	task number

GET Transcription Content

You can get the transcription content of the video through this interface, you need to provide the video number.

GET /api/serve/video/getTranscription

Request Example

import requests
headers = {"Authorization": "Bearer <YOUR_ACCESS_TOKEN>"} # access token
data={"taskNo":"<TASK_NO>"}
response = requests.get(
    "https://mavi-backend.memories.ai/api/serve/video/getTranscription",
    headers=headers,
    params=data
)
print(response.json())

Replace the YOUR_ACCESS_TOKEN in the above code as your access token, TASK_NO as your task number, you can get the transcription content of the video through the task number.

Request Parameters

Name	Location	Type	Required	Description
Authorization	header	string	Yes	authorization token
taskNo	query	string	Yes	task number

Response Example

If you provide a callback URL when submitting the task, a mesage containing content inside data in the following example will be sent to callback URL automatically to notify the status of the task.

Status code 200

{
  "code": "string",
  "msg": "string",
  "data": {
    "status": "FINISH",
    "type": "AUDIO",
    "videoNo": "videoNo_fd30e4c3700c",
    "taskNo": "taskNo_0a3f11298b9e",
    "transcriptions": [
      {
        "id": 0,
        "startTime": 0,
        "endTime": 0,
        "content": "content_0e5150607a47"
      }
    ]
  }
}

Name	Type	Required	Description
code	string	true	response code
msg	string	true	message with response code
data	object	true	JSON data
» taskNo	string	true	task number
» status	string	true	transcription status: FINISH, UNFINISHED
» type	string	true	transcription type
» videoNo	string	true	video number
» transcriptions	array	true	transcription content
»» id	number	true	transcription number
»» startTime	number	true	start time
»» endTime	number	true	end time
»» content	string	true	transcription content