TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Welcome to Waifu Labs v2: How Do AIs Create?

455 pointsby Cixelynover 3 years ago

26 comments

liuruover 3 years ago
Hey HN, one of the team members here!<p>I hope you all enjoy playing with the new and improved generator! We&#x27;ve been hard at work improving the model quality since the last time the site was posted[1]<p>As both a professional fantasy illustrator &amp; software engineer, I find the concept of AI creativity so fascinating. On one hand, I know that mathematically AI only can hallucinate images that fit within the distribution of things that it&#x27;s seen. But from the artist perspective, the model&#x27;s ability to blend two existing styles into something so distinctly new is so incredible (and not to mention also commercially useful!)<p>Anyways, happy to answer any question, thoughts, or concerns!<p>---<p>[1] <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=20511459" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=20511459</a>]
评论 #29903370 未加载
评论 #29901983 未加载
评论 #29904140 未加载
评论 #29903382 未加载
评论 #29901854 未加载
评论 #29902456 未加载
评论 #29902190 未加载
评论 #29902129 未加载
评论 #29904309 未加载
评论 #29904794 未加载
评论 #29901844 未加载
评论 #29914283 未加载
评论 #29904632 未加载
评论 #29902640 未加载
评论 #29906443 未加载
评论 #29904853 未加载
评论 #29923721 未加载
评论 #29901863 未加载
评论 #29903066 未加载
ve55over 3 years ago
Links to related projects in anime art generation for those interested:<p>Waifu Labs v2, referenced in this post (generate amazing custom anime face images): <a href="https:&#x2F;&#x2F;waifulabs.com" rel="nofollow">https:&#x2F;&#x2F;waifulabs.com</a> (write-up is the above link: <a href="https:&#x2F;&#x2F;waifulabs.com&#x2F;blog&#x2F;ai-creativity" rel="nofollow">https:&#x2F;&#x2F;waifulabs.com&#x2F;blog&#x2F;ai-creativity</a>)<p>This Anime Does Not Exist (AI-generated anime-style artwork): <a href="https:&#x2F;&#x2F;thisanimedoesnotexist.ai" rel="nofollow">https:&#x2F;&#x2F;thisanimedoesnotexist.ai</a> (write-up <a href="https:&#x2F;&#x2F;www.gwern.net&#x2F;Faces#extended-stylegan2-danbooru2019-aydao" rel="nofollow">https:&#x2F;&#x2F;www.gwern.net&#x2F;Faces#extended-stylegan2-danbooru2019-...</a> and <a href="https:&#x2F;&#x2F;nearcyan.com&#x2F;this-anime-does-not-exist" rel="nofollow">https:&#x2F;&#x2F;nearcyan.com&#x2F;this-anime-does-not-exist</a>)<p>This Waifu Does Not Exist (AI-generated anime-style faces): <a href="https:&#x2F;&#x2F;thiswaifudoesnotexist.net" rel="nofollow">https:&#x2F;&#x2F;thiswaifudoesnotexist.net</a> (write-up: <a href="https:&#x2F;&#x2F;www.gwern.net&#x2F;Faces#twdne" rel="nofollow">https:&#x2F;&#x2F;www.gwern.net&#x2F;Faces#twdne</a>)<p>There&#x27;s also a <i>lot</i> of literate on e.g. automatic manga coloration, auto-translation, image superresolution, anime frame interpolation, and much more. Worth checking out some places like <a href="https:&#x2F;&#x2F;old.reddit.com&#x2F;r&#x2F;AnimeResearch&#x2F;" rel="nofollow">https:&#x2F;&#x2F;old.reddit.com&#x2F;r&#x2F;AnimeResearch&#x2F;</a> if you&#x27;re interested!
评论 #29909421 未加载
评论 #29903703 未加载
Gigachadover 3 years ago
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.<p>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.<p>But the feeling this gives me, is what happens for the future of art. Sure, this example is no where even close to replacing real artists, but it’s already generating images better than I can draw after a year of practice. It does give me a feeling of “what is the point”. Which might be an irrational feeling, but I’m sure others feel the same.
评论 #29903024 未加载
评论 #29909519 未加载
评论 #29903157 未加载
评论 #29902854 未加载
评论 #29906806 未加载
评论 #29902718 未加载
评论 #29903542 未加载
bmitcover 3 years ago
This isn&#x27;t really creativity is it? I like to call stuff like this statistical copying, and indeed, the linked Wikipedia article on GANs says:<p>&gt; Given a training set, this technique learns to generate new data with the same statistics as the training set.<p>There isn&#x27;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&#x27;t any AI either. It&#x27;s machine learning, i.e., statistical models and algorithms.
评论 #29903579 未加载
评论 #29903282 未加载
评论 #29906906 未加载
评论 #29903331 未加载
评论 #29903400 未加载
评论 #29909369 未加载
评论 #29903668 未加载
deftover 3 years ago
I made a waifu, then I pressed save. It wanted me to signup so I pressed back so I could just rightclick saveas but... I lost my waifu forever now :(
评论 #29902775 未加载
评论 #29907246 未加载
评论 #29910053 未加载
评论 #29903795 未加载
评论 #29905853 未加载
评论 #29904976 未加载
评论 #29908497 未加载
评论 #29908591 未加载
tasha0663over 3 years ago
&gt; Step 43636: During this phase, the training gets unstable at times, so we have snapshots of occasional horrors like this.<p>Ah, make that <i>three</i> things the public shouldn&#x27;t see being made: sausage, legislation, and waifus.
jakey_bakeyover 3 years ago
I wish I had the stones to call my company Waifu Labs
评论 #29908267 未加载
mensetmanusmanover 3 years ago
I can’t wait for this technology to come to video.<p>Imagine a future where people can compile written scripts into Hollywood quality movies.
评论 #29902459 未加载
评论 #29902700 未加载
评论 #29902667 未加载
kadokaelanover 3 years ago
The explainer video about gans is top-notch! Excited for Arrowmancer!
评论 #29901874 未加载
评论 #29902147 未加载
评论 #29902458 未加载
echelonover 3 years ago
This is mind-blowingly good. You keep pushing the state of the art further to the point of broad applicability. It won&#x27;t be long until everyone can be an artist without putting in the ten thousand hours of drudgery of training their muscles, hand-eye coordination, structure of shape and perspective, etc. I can&#x27;t wait!<p>Do you have a team page? How many of you are there? Do you work with gwern and nearcyan? Are you going to raise for this? (You should totally scale this!)<p>Great work, and keep it up!
评论 #29914593 未加载
Kuinoxover 3 years ago
It&#x27;s funny how people complained about github copilot &#x27;stealing&#x27; people code, but nobody here complained about this AI &#x27;stealing&#x27; artworks.<p>Don&#x27;t get me wrong, I have nothing against this, but I think we should start discusing morality of AI generated content, even if it doesn&#x27;t train on existing artworks&#x2F;code.
评论 #29904667 未加载
syntheweaveover 3 years ago
The image quality is good, but now I realize I&#x27;m experiencing &quot;uncanny waifu&quot;. Authentic character designs bear two things in common:<p>1. Simplifications of reality(the actual artist training method would be traditional studies off life and photo reference followed by gradual reduction and symbolization to a style)<p>2. Symbolic meaning. Things like the style of eyes, clothing, etc are all meant to signal personality. This is stuff that current AI techniques don&#x27;t really touch upon in any direct sense.<p>Since the ML method is built on interpolating off final results, it&#x27;s going to lack in these qualities and produce something that is consistently an &quot;average impression&quot;. Akin to asking the algorithm to generate mythical heroes by mashing up the various stories: you get a hero that is somehow the average of Icarus, Heracles and Achilles, which would be less of a character than the originals.
评论 #29911578 未加载
w_t_payneover 3 years ago
I found the most interesting part was the evocative comment about the &#x27;vast and parched&#x27; nature of the latent space.<p>I wonder if the OP&#x27;s intuition regarding the sparseness of the latent space, and the relatively small area occupied by the &#x27;useful&#x27; manifold? embedded within it provide us any clues as to what symbol grounding might look like for some neuro-symbolic infrastructure that sits atop that latent space.<p>I.e. how should we be trying to represent concepts like &#x27;male&#x27; and &#x27;female&#x27; within that space?<p>Is it important to have these concepts represented as a low dimensional manifold?<p>Is it important that this manifold be easily described by some simple geometric form like a convex polytope?<p>Is it important that nuances and variations on the concept be separable within the bounds of the concept-specific manifold?<p>What other properties might be important?
aimorover 3 years ago
What methods are there to estimate how many unique characters a model can generate? The answer is not infinitely many, but determining when two images are of different &#x27;characters&#x27; is fuzzy.
评论 #29910507 未加载
dirtyidover 3 years ago
I hope these generators expands into non waifu &#x2F; pretty boy anime depictions. There&#x27;s a lot anime gaijin faces out there to explore.
评论 #29903307 未加载
评论 #29903762 未加载
dmixover 3 years ago
So putting <i>some</i> Anime digital illustrators out of work? ....or I could see many simply use them and pretend they did it themselves?<p>Obvioisly there will be plenty of illustrators doing custom work that these can&#x27;t (yet) replicate.<p>Also good for those countless anime avatar&#x27;d Twitter users.
Tade0over 3 years ago
Amazing work and progress. The previous version appears almost toyish in comparison.<p>FYI uBlock Origin complains about the registration link, because it on &quot;Peter Lowe’s Ad and tracking server list&quot;.
评论 #29905682 未加载
jcun4128over 3 years ago
I would be interested to find one I like and use it for a desktop companion project, not an original idea but I am not an artist ha.
spywaregorillaover 3 years ago
Is there any way to approximate embeddings for a novel image?<p>novel meaning user provided, not generated by the model or in the training set.
评论 #29902284 未加载
7373737373over 3 years ago
Have you considered applying similar models to VR avatar creation? That&#x27;s a market in itself
akomtuover 3 years ago
This tool has many applications, and those that will make you rich isn&#x27;t about anime.
评论 #29905467 未加载
评论 #29932376 未加载
评论 #29904032 未加载
ausbahover 3 years ago
any good info on manipulating the &quot;control vectors&quot; in the latent space?
easrngover 3 years ago
Cool, it&#x27;s like an AI-generated picrew
rambojohnsonover 3 years ago
project name is kinda incel cringe :(
ps91over 3 years ago
Amazing now I can create my own anime girl!
评论 #29902482 未加载
KaoruAoiShihoover 3 years ago
I don&#x27;t want to go all SJW on you guys, amazing work, but can you try to make sure there&#x27;s an inclusive array of starting faces please? Talking about things like skin tones, thanks!
评论 #29902439 未加载
评论 #29902448 未加载
评论 #29910549 未加载
评论 #29902413 未加载
评论 #29902350 未加载