As an aside, after some minor probing I haven’t seen much evidence that Bing/DE3 knows a style for Bluey. One set of images includes dogs, but it also includes boxes so I suspect that B/DE is trying to interpret the tokens as “Chewy”.
Star Wars in the style of Bluey
An angry anthropomorphic antelope in the style of Bluey
Frogs staging a production of Romeo and Juliet in the style of Bluey
Look for art generators that support ‘image to image’. There are some newer ones that will allow you to upload a bunch of pictures of someone, taken from different angles and different expressions, then create a model that allows you to put that person in any scene, in any pose, etc. There are even some out there that use ‘controlnet’ or similar tools to allow you to put a stick figure into a specific pose, then have the AI generate an image with your person in that pose.
Here’s one that seems to do what you want:
I just tried it, and it works pretty well. It only uses one photo, so not quite as accurate as some for some poses.
Playing around with Midjourney’s new style reference feature made me wonder if this isn’t intended in part to be an end run around some of the artists’ complaints and potential liability. Two artists that never showed up in MJ (using a prompt “Art by…”) were classic early fantasy/AD&D artists Erol Otus and David Trampier. With the style reference, I was able to point MJ at an image by them and, with a simple prompt, generate some art obviously using their influence.
Some of those are blended with other reference images since I wasn’t setting out to “make” an image by those artists but rather a thought I had after the fact. But I could see MJ saying “Oh, you don’t want your art in the model? Ok” knowing that anyone could just inject it back in on a per render basis anyway.
Revisiting “Girl with squirrel earring” (which was brought up in May of 22).
The squirrel didn’t have a right (screen left) rear leg and I shopped one in, and I outpainted the square inage, but otherwise it is as Bing created it.
Prompt was “Vermeer’s girl with a novelty squirrel earring. She has a grey squirrel on her ear.” and was my 6th try.
Mutant Ninja Dopers
Looks like some rabbit/turtle hanky-panky has taken place a generation back.
Resulting in a new cross-breed—turdits.
“Photo of anthropomorphic rabbit turtle hybrids wearing solid-colored t-shirts that say “mutant ninja dopers”.”
Are there any other Playground AI subscribers here? $15 dollars a month. For that you get the ability to create more images per day, more images at a time, and extra samplers. Here’s the problem … most of the samplers only seem to work about 50% of the time. They run fine for a while and then they start pumping out shit like this:
I don’t know if it’s me - a driver, because I’m using Firefox, I don’t know - but I’m getting kind of pissed off. I’ve complained to them via feedback a couple of times, but if they can’t get this addressed in some fashion, I’m unsubscribing. Too bad too, I like their interface. It’s fun to come up with different algorithms of filter upon filter and expanded text prompts.
That’s a great picture of a sailboat.
Fat lot you know - it’s a schooner.
Oh yeah!!! Well, THERE IS NO EASTER BUNNY!!!
Not in that picture, at least.
But seriously, that’s got to be a configuration issue on their end. I’ve got stuff like that running Stable Diffusion on my computer. It’s always a data flow problem.
I’m currently flooding my page with Magic Eye paintings with titles such as, “Playground AI Owes Me Fifteen Dollars!” and “Playground AI Are Crooks!” and “What do you have a feedback button for if you never address anything?”
I finally got some response from the site. Basically, “gee whiz, it does look like you’re having trouble with your images, we’ll have to look into that. Here’s $15 dollars back.”
So I unsubscribed.
I tried making a picture of a long-dead author in co-pilot, and it refised saying it violated their conditions. I just wanted a gray scale headshot of the aithor to go along with a post, but it wouldn’t do it. The last time I tried, it was fine, but wouldn’t let you reproduce faces of living people.
Someone I know tried to make an image with Gemini using a still from the Dire Strait’s “Money For Nothing” video and the AI refused, claiming that they were real people 
A couple of my recent favorite creations where I lora’d my face into them (from NightCafe):
Street Art Depicting the Map of a Serial Killer’s Soul:
A Priest with a Fist Full of Dollar Bills Watching a Nun Undress:
