Set a Focus Point

By default, the focus point of any asset used in Bynder is the center. You can change and customize the focus point if necessary. If you change the focus point, your cropped derivatives are updated on the fly, and the derivative is cropped from the focus point, not from the center.

If you are using Dynamic Asset Tranformation (DAT) and the system detects a face or group of faces, the face will be used as the focus point.

Learn how to set a focus point in DAT.

Note

  • You cannot set a focus point for watermarked assets.

This feature/solution requires Customer Support to enable, but then individual permissions can be done by the Bynder Admin.

Don't yet have Bynder? Start Here!

How to Set a Focus Point

  1. Navigate to your Asset Bank.
  2. Select an Asset  
  3. Navigate to the Asset detail page.
  4. Select the Set focus point tab.
  5. Change the values for the focus point. The new focus point is marked with a circle in the asset.

    Screen Shot 2016-11-28 at 18.19.07.png
  6. Click Save focus point. Your cropped derivatives will automatically update.

Note

The saved focus point is not respected for manual crops.

Automate Facial Detection

Automate Facial detection: If you are using Dynamic Asset Transformation (DAT), you can streamline your workflow by using our automated facial detection algorithms. The system identifies faces within your images and automatically sets the focus point on the subject rather than the default center of the image. While this detection typically occurs upon upload, you can now use the Override feature to apply or adjust detection on existing assets within the DAT UI.

Opt out & Reset: If you chose not to use the Automatic Focus point detector, you can chose to opt out your portal. After your portal has been opted out, you can request that the focus point is reset to the Gravity Center of the image (relative focus point of 0.5, 0.5). Please Contact your CSM to request and Opt out and Focus point Reset.

Disclaimer (Points of Interest): While the model is significantly better at identifying human subjects, a separate update for non-human "Points of Interest" (e.g., a product on a table or a specific architectural detail) is planned for a future improvement and will be released in a future phase.

Updated