Since last year I have trying out A.I artwork. It is a pretty powerful tool and can create a lot of beautiful images but it is also very random but I'm still learning how to us it properly. 
I tried a couple of different tools using Supergirl as my initial model. The ones in this set use a tool called Gliphy which sadly no longer exists in the form I was using it. One of the overwhelming aspects of AI is that it is constantly changing and growing so the images or techniques used to create content with it are constantly becoming obsolete. Glif is still currently available in App form but I haven't attempted using this new version of it.
When I was trying the earlier version. It was pretty good. With AI images the success rate especially if you are trying to create flawless images of barefoot women isn't particularly high. With Gliphy 20% of my attempts created usable images. I was able to up the success rate to 30% by editing some of the flawed images using Photoshop.
A lot of the prompts I was messing with with were very basic. Example: "Supergirl Barefoot on the beach during the sunset"
I played around variations of that such adding different scenarios i.e. "...she is playing with a puppy" or "her boots are on the side next to her."
There were some prompts additions that had really low success rate such as "She is showing her soles" or she is performing a kick where I could only get like one in 20 images to work with it.
This will be a regular series sharing my results trying different tools and characters. If you guys have any recommendations for tools to try creating A.I. Artwork with for future articles or prompts you would like to see just let me know.
Up up and away!
-KSC
No comments:
Post a Comment