Bringing comic characters to life with Artificial Intelligence.

What they really looked like! Creating photorealistic images from drawn cartoon characters.

I borrowed some cutting edge AI for a while and handed it some characters from my old commando comic compilations.

True Brit compilation in fact (if you want to buy it from Amazon). I’ve also put one of the stories in the slide show at the bottom of the page.

Anyway, I wanted to convert the images into photos, because AI can do that now.

“Hey, AI!” I said.

“Show me the real characters behind these drawings.”
I said

“Please!” I said.

There was no answer because it was a piece of software and so I had to prepare it myself.

I used gimp editing software to vaguely tidy up some images I’d photographed from the comic. I extracted some front facing images (because that’s a limitation at the moment) shoved them into the machine and waited for the results.

A couple of minutes later, there it was. The imaginary past brought to life. Real photograph-like images as if they were today’s brigands and heroes.

Gentle as pussycats they look. Don’t be fooled by those chiselled chins and thick monobrows, they were humans after all. Well . . .

The AI struggled a bit, because I didn’t have all front facing images and some had hats and things that it couldn’t handle but it produced something. They weren’t in my humble opinion what I would have imagined them to be like but here I am pretending I know better than AI. (For the sake of my future self I say here, “you did really well young AI wan.”)

Anyway, here are the first results.

This cutting edge science is moving so fast that it won’t be many years or months before we see huge improvements.

But, here are the few images I processed.

This was the original bit I cut out.

commando comic line drawing
commando comic line drawing

Here I roughly cut out the visage, and cleaned out the backgrounds a bit.

crop of character head, commando comic True Brit
crop of the head
cleaned up image
cleaned up a bit

I gave it to the machine and only then realised it’s limitations so some of the results are my fault.

and here’s the result . . .

composite photo realistic image from cartoon
here’s the result. A photorealistic image of a head from the cartoon. No sh*t!!

But wow! what a game changer. Just imagine what it’ll be able to do in a few years time. It’s so exciting ! ! !

Here are some more:

did the nose really well I think

The machine doesn’t quite know how to translate faces that aren’t straight on yet, and it doesn’t understand that eyes are usually the same colour, but OMG. This is going somewhere fast!