To clarify, I'm not the creator.
Sizigi Studios uses a dataset I put together (https://www.gwern.net/Danbooru2018) as their primary training corpus, AFAIK, and they were partially inspired by my application of GANs to anime art (as demonstrated on https://www.thiswaifudoesnotexist.net/ and see https://www.gwern.net/Faces for much more detail about every aspect of it), but I have never been involved with them and don't know much about what they've done other than what you can read in OP. It tickles me pink to see people following up on my anime GANs, though, especially as a startup! The Great Work goes on.
Re:How we built the Waifu Vending Machine
by kevinfrans on Tuesday July 23, 2019 @11:24PM
(#20512174)
One of the creators here -- gathering the Danbooru data in one place was definitely a big help! Anime's been a pretty nice space (both personally and research-wise) especially due to the abundance of data online, and it's great that we no longer have to manually scrape image hosts (which I've spent many hours doing in the past @ https://github.com/kvfrans/deepcolor, https://canvasdrawer.autodeskresearch.com/ etc).
Gwern created a similar thing, but I believe it was a different team who made the Waifu Vending Machine. Not least because there are pictures, and Gwern is quite known for their pseudonymity.
The steps could be something like this: (1) Select character, (2) Select colors, (3) Select outfit, (4) Select pose
The reason I suggest this is I was a little confused on what the options were, what my future options would be, and how many options are left.