Home Artists Posts Import Register

Content

I've been trying something new this week! Artificial Intelligence (or Machine Learning) programs that can create an image based on text input. This is *very* different from my usual process and I wanted to test ideas and techniques to see what I could do with it.

This is a gallery of "Near Misses"

All of these renders have some quirk or other, (like eyes with different colors or squished fingers, or something) but I thought you would like to see what I've been up to this week. The last two especially looked very promising, until I took a closer look. Yikes!

It is far too difficult to get the program to create exactly what I want it to - and it creates some really strange images at times (some can even be disturbing, like hands for feet or a tangle of arms and legs that look like two bodies sewn together and with no heads). Anyway, I will continue to explore what I can do with this technology and how I might integrate it into my workflow to enhance my artwork. I'm sure as I practice I will get better at telling the algorithm what I want (and what I do NOT want)!

Files

Comments

S Munson

Bravo for learning!!!

pratev

Bravo for machine learning! (Pratev learning takes *much* longer...)

pratev

Interesting, but not practical for more than supplemental use, I think. I spent *way* too much time messing with this stuff all week!

pratev

Thank you, J. I do agree, and I'm far from done with this new tool - I just got very distracted is all. My last render of the month went out in less than 4 hours total, including rendering time, and I was working on my little AI projects at the same time! I had fun, Steven (and am still having fun) !