Photoshop AI can instantly add elements and expand backgrounds

Adobe used the example of replacing lasso with noodles to show how quickly new AI tools can work.

  • Adobe debuted the AI ​​creation tool in Photoshop.
  • Generative fills can instantly add or remove elements from an image, or stretch an image from a text prompt.
  • The tool is currently available to paying Photoshop users in beta, and will be released to the public later this year.

Adobe Photoshop joins the generative AI boom sparked by OpenAI’s ChatGPT.

The company announced the new “Generative Fill” tool in a blog post on Tuesday. The tool can change an image’s sky from night to day, or as Adobe pointed out, you can replace your lasso with some noodles with a text prompt.

The company said the tool is now available to Photoshop subscribers in beta. The blog post outlines the basics of its new suite of AI tools, called “Firefly,” along with more than 20 sample images of how the Generative Fill tool can quickly add or remove details from images.

New AI features in Photoshop can make real-time adjustments to a basic image such as adding or removing features or changing the appearance of the sky.
Russell Preston Brown / Adobe

The new tools will allow Photoshop users to complete tasks that previously could have taken hours for even advanced photo editors in seconds, according to the post. As a play on the “Spaghetti Western” movie genre, Adobe showed off a simple version of what Generative Fill can do, showing how a cowboy lasso can be replaced with similarly shaped pasta.

Pam Clark, vice president of Photoshop product management and product strategy, said in a blog post that Firefly is Adobe’s most successful beta test ever, and said users have produced “more than 100 million assets” in the six weeks since the AI ​​tools were launched. She said providing AI image creation and editing capabilities in Photoshop is an “incredible new capability” for creators.

New Photoshop tools can add, remove, or replace a specific element in an image, such as adjustments made to the image of stones in the lawn above, but they can also expand the image based on a text prompt provided by the user:

New AI tools in Photoshop can expand an image based on what’s already in the image and a text prompt from the user.
New tools can take an image of a single flower and create a larger landscape around it.
Cory Schubert / Adobe

While the new tools aren’t meant for commercial use while in beta, Adobe said images should be free of copyright issues once the Generative Fill features are released to all Photoshop users by the end of the year. Adobe’s generative AI is trained on Adobe’s library of stock images, as well as other public domain and licensed images.

The popularity and quality of AI image generation tools along with generative script chat programs like ChatGPT have risen in recent months, but disputes over the copyrighted images being used have also led to lawsuits.

The edits Generative Fill does aren’t always perfect, as shown by examples created by a writer for The Verge. But as with other photo editing services like the one Google announced earlier this month, the technology is designed to learn and improve over time. Google recently announced similar features and other AI upgrades at its I/O developer conference, including the ability to expand existing images with new AI-generated parts, which are set to open to select users later this year.

Take a look at some other examples of what Photoshop’s new generative fill tool is capable of:

Generative Fill in Photoshop can turn a normal image of a lake into an expanded landscape with fancy buildings and a mountain that looks a bit out of place
Pam Clark / Adobe
Have you ever wanted to make your dog fly?
New AI tools can quickly change the sky or add individual objects such as a tent, a person, a shooting star or a flying car

#Photoshop #instantly #add #elements #expand #backgrounds

Leave a Comment

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

Scroll to Top