Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Welcome to Waifu Labs v2: How Do AIs Create?

waifulabs.com

Post reply on HN
Print view View original

71–80 of 236 posts

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#71

Post by numpad0 »

liuru wrote: post #35
KaoruAoiShiho wrote: post #30

I don't want to go all SJW on you guys, amazing work, but can you try to make sure there's an inclusive array of starting faces please? Talking about things like skin tones, thanks!

Indeed, we spent 2 years working on this!!! It's an extremely hard research problem, because darker skin tones account for only about 0.3% of all anime art produced in the world. We have employed an absolutely exhaustive array of art and data science tricks to give the model the ability to draw darker skin tones, though they are underrepresented. The results that you see today are the culmination of many months of ca…

I don't quite understand the need to mandate racial caricaturizations in every modes of communication. I would find it upsetting if every pieces of texts and quotes were prefaced with an "origin" indicator to aid forming prejudices. Why have it in manga?

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#72

Post by Gigachad »

Bombthecat wrote: post #65
unobatbayar wrote: post #63

Earlier quoted context omitted.

The quality and style is mindblowing! What data did you train?

The more interesting question: What about the sources for the AI to train? How are those artists paid? Do we need to pay them? Or if it used by an AI as train data, we just say: Its like a human learning?

There is nothing that suggests it should be any different than human learning legally. As long as the output is significantly unique, it shouldn’t have any copyright issues.

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#73

Post by Gigachad »

bmitc wrote: post #68
Gigachad wrote: post #46

This is extremely impressive. It’s the first GAN I have seen which lets you tweak the result in a meaningful way rather than being just random. I think the speed that GANs have come in to the world has really shaken people up and it’s hard to process what this all means and what it will result in. Especially the ones which generate based on real people. But the feeling this gives me, is what happens for the future of…

Art isn't about statical copying or drawing well. It's about human expression.

These generated pieces look extremely similar to real human art so they look very expressive to me. And you can even add your own input by hitting refresh until you get something similar to what you wanted.

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#74

Post by bmitc »

This isn't really creativity is it? I like to call stuff like this statistical copying, and indeed, the linked Wikipedia article on GANs says:

> Given a training set, this technique learns to generate new data with the same statistics as the training set.

There isn't a creative process here nor any creative introspection going on. While the technical results are impressive, this article does not address creativity even superficially, and just slaps the label on. There isn't any AI either. It's machine learning, i.e., statistical models and algorithms.

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#75

Post by Cixelyn »

Bombthecat wrote: post #67
Cixelyn wrote: post #9

Earlier quoted context omitted.

(link for the lazy: https://youtu.be/Pab8pG5WbXQ ) Thanks so much! It's done by our fantastic animator[1]! GANs are quite interesting and we didn't see many approachable explainer videos targeted at lay people, so we decided to make one ourselves! [1] https://twitter.com/bumblingbeebo

Is there a gameplay video? Or at least screenshots of arrowverse?

Here's our current game trailer! https://www.youtube.com/watch?v=8WvRgb6kh4s

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#76

Post by aniforprez »

bmitc wrote: post #74

This isn't really creativity is it? I like to call stuff like this statistical copying, and indeed, the linked Wikipedia article on GANs says: > Given a training set, this technique learns to generate new data with the same statistics as the training set. There isn't a creative process here nor any creative introspection going on. While the technical results are impressive, this article does not address creativity ev…

"create" doesn't necessarily imply "creativity" so I'm not sure why this is brought up. Of course generated images right now don't have any creativity attached to them and that's not the purpose of this article. All the images shown seem to mimic the generic "anime" art style and seems close enough to what a human draws which seems to have been the goal

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#77

Post by numpad0 »

greenn wrote: post #36
KaoruAoiShiho wrote: post #30

I don't want to go all SJW on you guys, amazing work, but can you try to make sure there's an inclusive array of starting faces please? Talking about things like skin tones, thanks!

Honestly asking, isn't this stuff usually Japanese characters? I dunno if theres even "data" they could use for other skin tones

Anything irrelevant to plot or story building, such as nationality, are usually left unspecified. Readers often project their own identity into characters, and some people seem to find it odd that they do not encounter traits that differ from their own.

Ethnicity is sometimes incorporated, that is, some distinctions would be necessary if there was a documentary manga about a match in an Olympic Games played by teams from multiple parts of the world, and in that case an American players might be given smaller eyes or extra wrinkles in face, or African players might be colored darker than other characters, Chinese players could be drawn with slightly different shapes of chins, etc.

But the default is unspecified or an averaged, most simplified shapes and forms that the author uses in their own cognition.

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#78

Post by aniforprez »

dirtyid wrote: post #62

I hope these generators expands into non waifu / pretty boy anime depictions. There's a lot anime gaijin faces out there to explore.

I would like to see how it generates late-80s/early-90s style features. The current pool of anime art styles are very generic (aside stylistic outliers) and I'd love to see Cowboy Bebop/Akira/Bubblegum Crisis type character designs

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#79

Post by numlock86 »

jhanschoo wrote: post #69
numlock86 wrote: post #51

Earlier quoted context omitted.

Well, there is (or rather was) "ganguro" at least, which appeared in mangas an animes occasionally, too. Not sure how many people would be getting offended by that these days, though.

Ganguro is associated with a certain Japanese subculture and aesthetic, and not meant at all to represent darker-skinned races. There are works that feature a more diverse cast, but perhaps the proportion of representation accurately reflects what you would see in Japanese society.

> Ganguro is associated with a certain Japanese subculture and aesthetic, [...]

Yes. What's your point?

> [...] and not meant at all to represent darker-skinned races.

No one ever claimed that. The question to my post was simply where to possibly get data from for "darker skin tones". Ganguro seems like an option because ... well, darker skin tones? Care to elaborate why you bring up "representation of darker skinned races" up in this context?

Re: Welcome to Waifu Labs v2: How Do AIs Create?

#80

Post by zaik »

bmitc wrote: post #74

This isn't really creativity is it? I like to call stuff like this statistical copying, and indeed, the linked Wikipedia article on GANs says: > Given a training set, this technique learns to generate new data with the same statistics as the training set. There isn't a creative process here nor any creative introspection going on. While the technical results are impressive, this article does not address creativity ev…

> There isn't any AI either. It's machine learning, i.e., statistical models and algorithms.

I always thaught those were synonyms.

Post reply on HN