> Yes. Image input is an anatomical skull. I'm currently using SD locally to train it on images in a slow attempt to lay the foundation for assistance in forensic reconstruction, and it ended up generating this. The skull that I used for this is from a 45 year old Chinese male. My hope is to be able to produce a fairly accurate "reconstruction" generator that will be cost effective and time effective with a minimum of 70% accurate likeness. This is a solo project because everyone I've pitched it to don't think it will work, but I know this is going to be the future, and I want to put this tech to good use and see what it can be capable of. I hope this works, to some effect because I have been looking at nothing but gore, forensic remains, cadavers and anatomical images for the last two weeks. My eyes need a break

OK. That's actually quite cool. Gross. But def cool