Unrecognizable face crop now has its own automation UI. Instead of using the Smart Crop automation type, we have carved out a new automation type to make it as simple as possible for creating AI-powered unrecognizable face crops. Learn more.

New automation type - Smart Resize + Background Remove - is now available. It works with both on-model and off-model photography. For instance, you can use it to remove or replace background on home furnishing products such as chairs! Supports person, products, car, animals and more. Learn more.

Smart Crop automation is now re-labelled as Custom Smart Crop. All the features such as crop around a person are still available. For creating unrecognizable face crop please use the new automation mentioned above. Learn more.

Background removal now works on images up to 25MB in size. The image pixel count doesn't matter. You can use 48MP images but they must be smaller than 25MB in size!

The crop mask handles at corners of a preview image are no longer cut off when outside the image canvas.

When downloading multiple images, the Zip generation no longer throws an exception. Previously users would see this error below that has now been resolved: java.lang.IllegalArgumentException: No valid customer id found for tenant id

Performance across the board for all automations will now see an improvement by at-least 20%. We continue to invest a lot of our energy in speeding up the automations.

Input images you upload can be of larger pixel dimensions such as 48MP but the output sizes you select must be less than 25MP. For example an output size of 5000x5000 pixels will be 25MP, this will work.
​
But if you choose a larger output size such as 6000x6000 pixels (30MP) that is not supported if background removal is selected.