Playground A.I. is another A.I. Engine with some limited free to use features that I was attempting to try last year. Of the three engines I was using, I found Playground didn't produce the best results for me compared to the other ones. It did have some interesting features though and there was some good material that came out of it. Anyway... Let's dive into this...
Playground A.I.
Please not the interface has updated a bit since the last time I was using it last summer. The way it was setup and performance of the engine may differ from the experience shared below.
The way that Playground is setup is that you select from Templates available and then select styles to giving different prompts to get the image you want
I selected art from the above menu and one of the starting images setup was the picture of Wonder Woman below. It was an easy pic to test with.
From there you could give different prompts to adjust the pictures or select from style lists (see below) to get the image you want.
You also the option to undo a change if you didn't like the outcome un update you made to the image. With barefoot characters there are some bizarre outcomes so the undo was really helpful. I played around with it trying a variety of prompts and styles. I gave instructions such as:
"Make the image Supergirl instead"
"Open her eyes"
"Change the time of day to sunset"
And of course...
"Make her barefoot"
Not all the prompts worked. 😅 For example "have her show the soles of her feet was a pretty big fail"
I also tried a variety of styles I liked some of the outcomes like Pastel provided some cute images:
Playground also has a cool feature that allows you to upload your own image as a starting point. I picked a classic image of Power Girl showing off her soles to test it
Playground actually fairly well though not without flaws.
I tried again this time removing the comic book text from the image. Pro-tip... It is good to have an image editing tool like Photoshop to assist with A.I. Generation as well in case you need to fix anything.
From there I played around with the prompt a bit asking it to update the subject to Supergirl and ran it through some of the styles available. There were some misses but these were some the better images it generated below:
I also tired uploading an image of a girl in the pose as a starting image to get the image below. Though I don't remember which image I used as reference.
It sort of looks like Melissa Benoist so images of her must be part of the training model for Supergirl.
Ok that will be it for today. If you guys have any recommendations of free A.I. tools to use or any character or poses that you would like me to play around with in future posts let me know.
For more Supergirl please check the link below:
and for more adventures in A.I. check out this earlier post:
And more of both of these categories to come in the future. ^_^
Up up and away!!
-KSC