That game looks amazing!
I prefer pixel-art for the nostalgia/vibe, but I’d probably give more high-res games a chance if they looked like that.
@seguso Impressive results, but as you haven’t given any hints in your post, could you explain what you are doing? Are you recreating each scene from Monkey Island or are you trying to create a whole adventure? Are you using/trying different settings in the AI? What are the differences between the pictures/posts - just the input picture?
I’m just posting random results in case someone here finds them useful.
I’m also ready to make specific backgrounds when and if we come up with specific requirements , about location and objects contained and puzzles.
a degree of manual retouching will be needed, to erase AI artifacts and to add specific objects. I’m ready to make them when needed.— after we have a story.
the last post contains 3 pictures generated from the same input picture, with different random seed
. And a big degree of freedom for the AI.
I just realized - we also need box art for big box releases.
@Paul and @seguso: Are you working with (a current version of) Photoshop? If you do, have you tried Adobe Firefly? It’s able to “fill” parts of a picture with objects generates by an AI.
Wow, sounds useful. Can you specify what object to generate?
Leonardo AI has something similar, in the “canvas” mode, but I didn’t manage to make it work yet.
Yes. As far as I understood, you select a region in your picture, call Firefly, enter a description of the object (as a text, like in other AI tools), select one of the generated objects and Photoshop paste it in your picture.
Please note that I haven’t tried it (yet), because I haven’t a subscription of Photoshop.
I don’t have Photoshop (I have Affinity Photo, as it was cheap, haha), but that sounds cool… I wonder if it blends the new stuff seamlessly into the existing background.
it’s free , and it works well
You can add objects easily with firefly: I added the two statues, which weren’t there in the original pic:
(and the chandelier, and the inscription on the left wall)
it’s free, web based, you don’t need photoshop. here’s how it looks:
Amazing stuff, also will be good for correcting “mistakes” that the original AI might have generated
When I try to download an image I’ve altered, it puts a big tag in the corner saying “not for commercial use”
Starring six-fingered, smug looking Gushbruy Fleepwoot, no doubt
. (The art looks kinda cool, actually. I like the inked and colored style.)
I’m missing the cows in your box art…
Take a screenshot! Problem solved ![]()
I figured out how to do generative fill on Leonardo AI, I watched this video on it -
The first half explains expanding a picture, and the second half explains erasing and filling in the space with a prompt (it’s similar to Firefly).
If it uses your trained model, it should be better than Firefly. We’ll see…





