Adobe Brings Firefly Generative AI Technology to Lightroom

Adobe Brings Firefly Generative AI Technology to Lightroom

Today, Adobe brings its popular Firefly Generative AI technology to Lightroom in the form of Generative Remove and Lens Blur. These features will be available in all Lightroom incarnations, including the mobile app, desktop app, and web browser versions.

For me, the most exciting news is the addition of Generative Remove in Lightroom, available today as an early access feature. This feature makes it easy to remove undesirable elements from a photograph. Generative Remove can eliminate minor elements such as a stray hair on the face or more complex distractions like an unwanted person standing in the background of an environmental portrait. Accessing Generative Remove is no more complicated than cropping a photograph or adding a preset to an image. Using a brush tool, the user can make a quick selection of the unwanted element, and Lightroom will present three versions of the photograph with that element removed. The user can select which version they prefer.

As a headshot photographer, I use Adobe’s Generative Fill in Photoshop to straighten ties, remove stray hairs, smooth away wrinkles in clothing, and remove skin imperfections. In my current workflow, however, the image must be exported to Photoshop for these adjustments. The new workflow allows for these AI adjustments to be made directly in Lightroom.

Although the feature currently uses Adobe’s older Image 1 Model rather than the newer Image 3 Model available on the company’s Firefly platform, Generative Remove is superior to the Healing Brush, which performs a similar function. Using the Healing Brush, the user must select an area that the program will use to replace the undesired element. Imagine a seamless paper backdrop that is torn on one side. The user would select an area from the undamaged section of the backdrop that the program will use to cover the undesired element. This often works well, but there are times when there is no area in the photograph that can be used to cover the unwanted element. Generative Remove works differently in that the program understands the composition of the photograph, and Lightroom creates what the program believes would have been shown in the photograph if the unwanted element had not been present.

The feature is easy to use, and photographers are likely to utilize it not only for perfecting their professional client work but for improving photographs taken for personal use as well. Lightroom attempts to retain the color tone, detail, and grain of the original image.

Lens Blur is also available today. The AI-powered feature takes into account the distance and design of background elements to ensure that the added blur effect is realistic and is as close as possible to what the photograph might have looked like if it had been created with camera settings that allowed for a shallow depth of field. Lens Blur also includes presets that the user can adjust to personal taste. Lens Blur can be used in conjunction with Generative Remove. Imagine a portrait taken on a busy street. The user might use Lens Blur to make the background elements more fuzzy and less distracting to the viewer. After doing so, the user sees that there is still a distracting figure in the background. Using Generative Remove, that person can be removed from the image. Once the person has been replaced, the user can use the blur tool to modify the sharpness or fuzziness of the section of the image that has been added through AI to ensure all parts of the image look organic to the final edit of the photograph.

According to Adobe, a recent study conducted by the company indicates that 76% of U.S. consumers emphasized the importance of knowing if online content is generated using AI. For that reason, Content Credentials are automatically attached to photos edited with Generative Remove and Lens Blur. Content Credentials are tamper-evident metadata that provides viewers with information about the creation, modification, and publication of images and artwork. This information allows for transparency in the creative process so that a viewer knows if an image has been adjusted. Likewise, a viewer can use Content Credentials to ensure that an image accompanying a news article has not been modified after its initial capture. More information about Content Credentials can be found at the Content Authenticity Initiative (CAI) website. 

In other news announced today, Adobe will extend its enterprise supply chain solution with new AI-powered capabilities in Adobe Express for Enterprise. In doing so, Adobe will be greatly increasing the number of people who can create content and publish across any channel, including:

  • Social media teams with an ongoing need for new content
  • Team members who don’t have budgets to outsource the creation of content
  • Sales teams that need to constantly update decks to customize content for pitches
  • Comms teams that need to create decks for media briefings or reports for internal stakeholders

Adobe is quick to point out that over 8 billion images have been created or modified with its generative AI features since its introduction in March 2023. Firefly is trained on content licensed by Adobe, and any images created or modified in Lightroom will not infringe on copyright and other intellectual property (IP) rights such as likenesses, locations, trademarks, or logos. As a result, companies using Adobe Express for Enterprise can allow any team members to utilize Adobe's AI technology for content creation or modification without securing licensing or usage permissions.

John Ricard's picture

John Ricard is a NYC based portrait photographer. You can find more of Ricard’s work on his Instagram. accounts, www.instagram.com/JohnRicard and www.instagram.com/RicInAction

Log in or register to post comments
3 Comments

I like all the new AI features and everything, but my favorite new update is that they FINALLY added tethering for the Canon R6 II.

This will prove useful to quickly replace Arne Slott with Jurgen Klopp in next season's photos of the Liverpool touchline.

It's still not up to PS standards.I had an image with 2 motorcycles in it and used the select subject feature in LRC and it selected only one bike and did a crappy job at that, then did the same thing in Photo Shop Beta and it selected both bikes and did a good job including picking out only the spokes and not the background between the spokes.