Lightroom tip: How to get Lightroom’s new Distraction Removal (People) tool to work!

This post was written soon after the new Distraction tools were added to Lightroom, so some of the issues I talk about may have been solved by the time you read this. This post is for anyone struggling to get the early version working…

Adobe’s new AI-powered Distraction Removal tools are accessed via the Remove tool. This needs to be set to the Remove mode (not Heal or Clone) and with the Generative AI box checked. I also checked the Detect Objects box.

Now the problem I had was that the two Distraction Removal options – Reflections and People – were greyed out. I couldn’t work out why until Google found an answer. Someone suggested making an image adjustment, e.g. Exposure, and then trying again. Sure enough, that works! It also works if you immediately undo your adjustments. It just seems to ‘wake up’ the Distraction Removal options so that they become available. Obviously a bug, but this is a workaround until Adobe fixes it.

In this post I’ll look at the People option and I’ll cover Reflection removal in a separate post. They are all lumped together in the interface, but they are quite different.

Distractions > People: How does it work, is it any good?

When you click the disclosure arrow next to the People section, Lightroom will check the image for distracting people and, if it finds any people it thinks it can remove, it highlights them with a red mask overlay. You don’t have any control over what Lightroom initially identifies as people that can be removed, but each one has its own handle which you can select to delete it.

All that’s left to do now is click the Remove button and wait for Lightroom’s object removal AI to do its work. With some images it does a great job, with others it’s clearly struggling and will insert areas at a lower resolution than the original image – a persistent issue with Adobe’s generative AI. Occasionally it will mess up completely and fill the repaired areas with half-objects, weird shapes and other AI guesswork that’s just unusable.

You do have a degree of control after the event, so that you can click the control handles for each object removal area then choose one of three different variants. It’s definitely worth having and might fix an issue with the result, but you may still end up with odd AI artefacts and inadequate resolution.

So this is a neat tool to have, but as with so much AI it either works or it doesn’t, and if it doesn’t work then there’s no a whole lot you can do about it. The artefacts and resolution limits also mean this is perhaps fine for social posts but not for high-quality work. I might try it out now and again to see if Adobe has improved it, but it’s not going to be part of my daily edits.

In fairness, this is still an early access tool that’s still being developed, and it does have a Provide Feedback button for anyone who wants to help it along a little.