A.I., Machine Learning, and zombies

Ever since I got my first Commodore Vic20 (and then later the Commodore 64), I’ve been addicted to using the computer to make art. And while my college years were spent studying traditional painting and drawing, as soon as I graduated and moved to Los Angeles, I dove headfirst into using the computer to animate and paint.

Recently, the field of A.I and Machine Learning has made leaps and bounds, making these tools finally available to artists. I first started playing around with a web-based app called Pix2Pix, which let you sketch something, and the computer would “flesh” it out.

My initial results were pure nightmare fuel.

Not too long after, I discovered a site called Artbreeder, which gave access to a pre-trained set of models. Mostly you could make your photos look like Van Gogh, or any other artist whose work the A.I. had been trained on. But I wanted to be able to train my own model. I wanted the computer to make new work based on my entire body of digital paintings.

That’s when I found RunwayML.

RunwayML’s aim was to make the training of new A.I. models accessible to artists, which is exactly what I was wishing for. And their site made the whole process super painless to hop into. Soon I was training models on my amoeba characters, some digital faces, and now, zombies.

My first experiment in the land of the A.I. undead, was to grab all the photos of zombies I could find from The Walking Dead. I trained a model on them, and it started giving me some pretty creepy, but awesome output. Soon, I realized that I would have to fix some of its attempts, and re-train it to include the newly altered imagery.

This idea turned to work extremely well.

Then I decided to start re-painting the images to be a little sillier, more in my style. Those results were just as weird.

But a few days ago, I realized I had enough imagery of my silly, squiggly zombies to try training a model on those. And boy am I glad I did! Right away the computer started making things that approximated my wide-eyed, confused zombies. I just had to keep training it.

But then something went wrong. In Machine Learning, there is a measurement used to describe how confident the computer is in its attempts to recreate the source imagery. This is called the FID score, or Frechet Inception Distance. It’s basically a measure of how far off the pixels of the generated image are from the original images in the training data-set.

And somewhere in my over-excitement, I had over-trained my zombie model.

So my current goal, is to take the output from the stuff that worked, and clean it up. Re-paint it all and feed it back to the computer. Like I did with the Walking Dead zombies. My hope is that with a little time and care, I can train a little robot Petey to make very good silly squiggly zombies. Just like his dad.

Lunchtime beat-making and Mayan art

Lately I’ve been playing around a lot with the Pocket Operator PO-33. It’s a fantastic pocket-sized sampler and step-sequencer, which means you can make fun little beats out of almost anything. Which has become one of my favorite things to do.

My new habit is to take it to lunch with me and sample the surroundings, and then turn that into a little looped beat. My most recent lunchtime beat-making adventure was at In-n-Out.

After lunch, I took it to a nearby park, and made some drum sounds by banging on things. I found a metal pole that has some great resonance, like an old bell, so I sampled that. It became the basis for another lunchtime beat.

Then I made some new art inspired by Mayan and Aztec culture as part of my series Lightbearer.

View fullsize

I have always been fascinated by Mayan and Aztec art. There's such an interesting otherworldly feel to the stuff. A while ago I started a project called Lightbearer, which was inspired by the art of those cultures. The idea is that there are keepers of the Light. Guardians who protect it from the Beasts. The Beasts want nothing more than to steal the Light.

This week, I made a bunch of new images for this series, inspired by Mayan and Aztec art. Check them out in my shop.

https://www.etsy.com/shop/daspetey

Poured Painting

Recently I discovered a new painting technique that I immediately fell in love with. It's called poured painting. The idea is to mix acrylic paint with some kind of flow medium, Floetrol or Liquitex flow medium, to thin the paint out. Then using a variety of techniques you simply pour the paint onto the canvas, and tilt it around to get even coverage.

But the real fun happens when you add an oil to the mix. A few drops of silicone mixed into each color keeps them from fully blending with the other colors. It helps them start to form interesting cell patterns and swirls.

I'm definitely in the early stages of my knowledge with this technique and I learn a lot with each new painting I do. But the process is fun and medidative, and I really love the results.

You can see a little more of this in my newest Vlog.

Draw Zombies with me!

Not too long ago I made a Skillshare class on drawing silly little zombies. The class is pretty short, and pretty silly. But best of all, it's free!

Years ago, before I got signed to a record label and went on tour, I was teaching animation at a small tech college in L.A.  It was super fun, and part of me misses that. So I think a big reason I wanted to make a Skillshare class was so I could quench that thirst again.

The trick was figuring out how to break down the process. How was I going to walk students through my oddball method for making tiny little undead hordes. I brainstormed a few possibilities and ultimately settled on what became the version of the class that's on Skillshare now.
 

In the class, you'll learn to make your very own hungry zombie horde.

I knew even as I was filming the zombie class, that I'd soon want to make more classes. Some more drawing ones of course, but even some music ones. I've developed some fun ways of coming up with song ideas over the years, and I think it would be fun to share some of those methods.

But for now, let's just draw some zombies!
-pete

Diamondhead Covers

I'm having trouble deciding which cover image to use.

Option 1

Option 2

Option 3

Option 4

Option 5