Start by selecting one or more file from your local system to upload to Veritone. Once your file is selected, you’ll choose engines to create a transcription job with speaker separation. 

Upload a File

1. Click New and select Upload from the dropdown. The File Upload tab opens to the Upload page.

2. Click Upload Media in the middle of the window. The File Picker window opens.

3. Drag and drop a file from your desktop or click Choose File to upload one or more files from your local file system. The selected files display in the File Picker window.

4. Click Upload. The file(s) display on the Upload Media page.

5. Add optional configurations to a file, such as a cover image or tags. (You may skip this step and proceed to #8 below.) Click a file to select it and click the green edit icon. The Edit Media dialog box opens. 

6. Add details to identify your file.

   a. Upload an avatar image. 

   b. Upload a cover image.

   c. Change the file name. 

   d. Add tags to make it easier to find the file.

7. Click Save when all configurations are added. The Edit Media dialog box closes.

8. If more than one file is being uploaded, click Select All to upload all files in the list.

9. Click Next. The Processing tab opens to the Simple Cognitive Workflow.

Create a Transcription Job with Speaker Separation

10. Click Show Advanced Cognitive Workflow in the upper right of the window. The Advanced Cognitive Workflow opens.

11. Click the Available Engines dropdown and select Transcription from the list. Available transcription engines displays. 

12. Click the + on the right of the transcription engine(s) you would like to use. (Tip: Hover over the engine name in the list to view details about the engine.) The selected engine displays in the Selected Engines box.

Note: If a “Missing Dependency” message displays, the file must be transcoded before it can be transcribed. To fix this, click the Available Engines dropdown and select Transcode. Then click the + on the right of a transcoding engine to add it to your request.

13. Enter additional information required by the engine for processing (if applicable).

14. Click the Available Engines dropdown and select Speaker Separation. Available speaker separation engines display. 

15. Click the + on the right of the speaker detection engine you would like to use. The selected engine displays in the Selected Engines box.

16. Click Next at the bottom right of the window. The Content Templates page opens.

17. No action is required on the Content Templates page. Click Next in the bottom right of the window. The Customize tab opens to the Ingestion Customization page.

18. By default, the file will be added to your organization’s private media collection in CMS. Choose to add the file to a specific folder, if desired.

19. Add optional tags to help find the file through search.

20. Click Save. The CMS homepage re-opens and a green “Your ingestion job is now active” message displays momentarily in the upper right of the window.

Depending on the size of the file being uploaded, it may take a few minutes before it appears on the CMS homepage. You can check the status of your transcription job using the steps described in Check the Status of a Transcription Job.

Did this answer your question?