Unleash live's annotation capability allows users to mark up areas of an image with a variety of context-specific labels (defined by the A.I. model applied). These annotations can be used to guide manual stakeholder review or used as the basis for improving an A.I. model (retraining).
Annotations are applied to an area of an image and can be set as a multi-point polygon or a rectangle.
The video below outlines the procedure to annotate an image:
- Go to Media Drive and open the image that you want to annotate.
- Navigate to the Annotation tab in the right sidebar.
- Choose the appropriate A.I. App from the drop-down menu of the Annotation Category. Note: The A.I. App must be added to your account from the A.I. App Store before it will be available in the drop-down menu.
- Click "+ ADD ANNOTATION" and select either the Polygon or Rectangle option.
- For Polygons:
- Select the first point using the left mouse click and then click at each point in the polygon. Once you have finished outlining the area, right-click to finalise the polygon or click on the starting point of the polygon. Note: individual points can be moved after the polygon has been finalised by left-clicking and dragging the point.
- For Rectangles:
- Left-click in the first corner of the rectangle, move the cursor to the opposite corner of the rectangle you desire and then left-click again to set the rectangular area. Again, the rectangle can be moved or resized after being set.
- Select the suitable annotation label and severity using the radio buttons and severity options available in the right sidebar.
- Add a comment in the free field area underneath, if needed
- Click on the Save button at the bottom of the sidebar.
After an image has been annotated, you are able to export it to CSV/JSON/PDF format using the Export Annotations option at the top of the page.
This article contains more information on exporting annotations.
For more information, please contact email@example.com and the support team will get back to you shortly.