Adobe Scene Stitch is content aware fill on steroids – Uses AI and image databases to fill holes

Oct 23, 2017

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Adobe Scene Stitch is content aware fill on steroids – Uses AI and image databases to fill holes

Oct 23, 2017

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Join the Discussion

Share on:

This is one of those features that many of us wish was added years ago. It’s common for photographers to fill in holes from one photograph, of things they want to remove, from sections of another. I’ve done it many times myself to fill in undergrowth in forests or even rebuild entire moss-covered walls behind my subject on location. But it’s a pain, and takes forever to do well.

Shown off at Adobe MAX, Adobe’s new Scene Stitch technology aims to solve this. Essentially, it works like Content Aware Fill. You select an area you wish to replace, and Photoshop figures out what to put in its place. Instead of looking only at the loaded image, though, it will look at a huge database of images. Then presents several possible options to fill the space.

YouTube video

In this case, the database used is the Adobe Stock library. The software’s AI searches for images that fit the current scene, and then uses areas from them to generate missing data.

It’s not just looking at image types, though. We know that Adobe, Facebook and other companies have AI that can detect “landscape”, “mountain”, “trees”, etc. But it’s actually looking to see what fits, aesthetically.

This scene is used as one example in the video. You can see with the use of the standard Content Aware Fill, that it doesn’t do a great job. It’s funny to compare how highly Adobe praised the Content Aware Fill feature vs. how they seem to deride it now.

And the results work to varying levels of success. Not all are perfect, but some you’d never know anything was changed.

As with the other tech sneak peeks offered at Adobe MAX, there’s no guarantee that this will show up in production software. Ever. But with something like this, I think it’s difficult to imagine it wouldn’t find its way into Photoshop at some point.

Hopefully, it will allow you to limit the source search to local images and not just Adobe’s stock library.

Filed Under:

Tagged With:

Find this interesting? Share it with your friends!

John Aldred

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Join the Discussion

DIYP Comment Policy
Be nice, be on-topic, no personal information or flames.

Leave a Reply

Your email address will not be published. Required fields are marked *