Page 1 of 1

Photoshop AI in 2024 - new possibilities

Posted: Sun Jan 26, 2025 8:40 am
by mostakimvip06
Adobe Photoshop beta has expanded the existing generative features, giving us the ability to, among other things, generate entire images or use reference images.
The beta version is available to every Adobe user and can be easily installed from within the Creative Cloud application.
The new generative features can also be used using the Adobe Firefly browser tool.
Contents
Previous AI features in Adobe Photoshop:
New AI features (currently available in beta - as of 10/05/2024)
How to install beta version?
Do I need to install Photoshop to use generative capabilities?
Exactly one year ago, in May 2023, our eyes were opened to incredible possibilities in the Photoshop application, namely generative fill . Since then, graphic work with images, especially photo manipulation or photomontage, has become much easier, faster and more enjoyable. Today, a year after the premiere, we have a new, major update of functions based on artificial intelligence, which is currently in beta, but will certainly be introduced to the standard version soon. In today's entry, we will take a look at the current possibilities of AI in Photoshop and discover new conveniences and functions. It is worth mentioning that we are working and talking about the version for desktop computers with Windows.

Previous AI features in Adobe Photoshop:
Generative filling
That is, the ability to generate graphic elements using artificial intelligence in a place we choose in the photo. The elements created can be based on the prompt we enter or, if we do not specify a prompt, on the surroundings of the selected fragment, i.e. the rest of the photo.

Photoshop AI generated poufDeleted image in Photoshop AIPhotoshop AI generated children element
Generative extension
This is the ability to crop a photo "upwards", i.e. instead of cutting the photo along its selected sides, we can add the photo where it was originally cropped when it was taken. Thanks to artificial intelligence, Photoshop will "add" a fragment of the photo in this place in such a way that it works with the entire composition, or if we specify a prompt - it will add the elements we choose there.

Photoshop AI Extended Background
New AI features (currently available in beta - as of 10/05/2024)
1. Generating whole images
With the Beta update, we gain the ability to create entire compositions, instead of just ivory coast phone numbers parts of them, as before. Generation is done by entering a prompt as standard, although from now on we will also have the ability to use additional controls, such as effects or specifying the content type.

Image generation area in Photoshop AI
Shark in a glass of water generated using Photoshop AI
Prompt: Small shark swimming in a glass of water standing on a table

Content type: photography

Effects: Hyperrealism

2. Generating reference image variants
This feature allows us to upload our own photo as a reference image, so we can generate images similar to our reference image.

AI generated image with reference to example
Prompt: Modern office with an aquarium with sharks

Reference image visible in the upper left corner

3. Generating similar images
Thanks to this feature, we can easily use generative AI to generate images similar to the ones we currently have in our document.

Similar graphics generated from the previous example:

Generated similar image in PhotoShop AI
4. Improved results of generative tools
With the latest update of Photoshop, we will receive generative functions based on the new version of the Adobe Firefly 3 model, which, compared to version 2, is supposed to bring us more satisfactory results of generated graphic elements of higher quality than before.

Adobe Firefly Logo
5. Generating the background
Thanks to the new features of Adobe Photoshop we also get the ability to change the background in our photo much easier and faster, after uploading it, just remove the background from the context bar and then go to the background generation option, where we can add a text prompt if necessary.

Background generated with Adobe Firefly
6. Enhancing image details
Until now, the elements generated by artificial intelligence were visibly of lower resolution. Thanks to the update, we have the ability to improve the details of the fragment of the photo generated by us. However, in order to use this function, the generated element should be small (low resolution), otherwise we will receive information that the selected variant is too small to be improved.

Example without correcting details:

AI enhanced image
After correcting the details:

AI sharpened image
It is true that when you enlarge the image significantly, and are aware that the photo has been edited, you can still see the differences. However, I assure you that for someone who sees the photo for the first time and who will not analyze it pixel by pixel - the transition between the original image and the creation will be imperceptible.

Improving the generation nevertheless gives a really good effect, as you can see in the screenshot below.