Imaginary worlds dreamed by BigGAN | Hacker News Reader

Imaginary worlds dreamed by BigGAN

aiweirdness.com

129 pointssytelus35 comments

You can now demo BigGAN in an official Colaboratory notebook (backed by a GPU) to create your own AI-generated nightmares: https://colab.research.google.com/github/tensorflow/hub/blob...

Examples from the notebook by the article author: https://twitter.com/JanelleCShane/status/1062067001504321536

Can someone explain to a layman what does the truncation parameter do?

My understanding is that the 'truncation' parameter makes the output of BigGAN more closely resemble the training data set.

Moving the truncation towards 0 will give you more realistic but less varied images, while moving it towards 1 will give you more varied but increasingly potentially nonsensical images.

Aha!

That makes sense. Thanks!

> Combine sugar and sugar and sprinkle with the sugar.

Seems about right describing current food lol

On a more serious note I wonder if our grand-grandchildren will look at these programs from blog posts and see them as some kind of "PRINT HELLO; GOTO 10" for AI, ancient artifacts showing the birth of a new paradigm of automation

> Combine sugar and sugar and sprinkle with the sugar.

What is this quote from?

An earlier entry from the same blog on generated recipes. My wife and I were laughing so hard at these we got some weird looks from the waiter. The use of asterisks is priceless in that one.

Strangely enough, on mobile all those posts were melted into one big page so I didn't even realize it was not the one linked to.

http://aiweirdness.com/post/176589646292/its-time-for-cookin... this apparently

Funny how they mess with your brain: the illusion is that you are looking at something familiar. The texture is "easy" to look at, but the brain fails to recognize a known object.

Interesting article, but the subject of the article (the amazing images) are only shown in tiny thumbnails, with no links to the full-size images. Seems to sort of defeat the purpose.

Even the largest images produced by the network are only 512px x 512px and most are smaller — adding more pixels to these kinds of networks is fairly expensive and not at all worth it if you're training thousands of them over and over again while doing research.

The full images are only 512x512.

One day this will generate graphics in games. Imagine Doom with synthetic monsters controlled by you biggest fear.

This was a Black Mirror episode.

Or No Man's Sky with actually infinitely many unique planets.

Or as I like to say, Skyrim with infinite content.

after it gets used by the Adult film industry.

We have millions of pornography to train BigGAN.

on a side note: GAN means "FUCK" in mandarin.

Read some other entries in that blog, recipes are something to look at: http://aiweirdness.com/post/176589646292/its-time-for-cookin...

More of the same: https://www.reddit.com/r/SubredditSimulator/

SubredditSimulator uses Markov chains to generate text, not an neural network.

I have made my own subreddit using the same RNN tool as that post: https://reddit.com/r/SubredditNN

have you tried partnering with simulator so your posts are displayed there too? I imagine they would get more exposure and therefore better feedback

I believe development on SubredditSimulator is stalled now that the dev is no longer working at Reddit / is building a Reddit competitor.

Makes it more likely for him to agree :)

The train picture reminds me of some of my botched attempts at creating perspective by combing two differing horizon line sub-images into one. I suppose that sort of rendering abstraction is the same in terms of accidentally producing weirdness - even though different processes were used.

How machines hallucinate.

https://twitter.com/quasimondo/status/1063449767617937410

We should have a reverse image search built in image-net for someone doubting and trying to find similar images in imagenet . Like google's reverse image search.

RIP stock photography. And possibly videography too.

Very doubtful, the results just seem to be combination of existing background and existing artifacts (nose, ears, furs, etc) .

The different poses you would get naturally is much more complicated and diverse than these networks seem to be generating. All in all these networks aren't generating anything new.

The images you saw in the article were hand picked to be weird. Most of them are photorealistic and are not copies of any training photo. If you think it's so easy, paint a photorealistic image in Photoshop by hand (no direct copying, but you can look at other images all you want) and let's see how it compares.

Cropped a few more BigGAN samples for reference: https://imgur.com/a/GIQSo4I

Or how about ProGAN for faces? https://www.youtube.com/watch?v=36lE9tV9vm0 Does that look like a simple copy? It's a 'walk in the latent space' of faces.

Also, take a look at this tool (TL-GAN) https://www.youtube.com/watch?v=O1by05eX424

That comes with an online demo: https://www.kaggle.com/summitkwan/tl-gan-demo

> photorealistic image in Photoshop by hand

If you are competing with a human, if not me I am sure a professional can easily better the best of best (and not just 512x512 pixels but 100 time more pixels) by this network. It has to catch up with humans by a long shot.

> Cropped a few more BigGAN samples for reference: https://imgur.com/a/GIQSo4I

I don't know but to me all of the images you have in the imgur link seem non realistic. Head or torso or legs is messed up for (all?)most of them.

> how about ProGAN for faces? https://www.youtube.com/watch?v=36lE9tV9vm0 Does that look like a simple copy? It's a 'walk in the latent space' of faces.

No, I am not saying its a simple copy, the copy is great. The fact is that its a copy. An imperfect one.

> Also, take a look at this tool (TL-GAN) https://www.youtube.com/watch?v=O1by05eX424

Sure, human face comes in all shapes and sizes, so its tricky to judge the effectiveness since every modification is a possible real person.

> Sure, human face comes in all shapes and sizes, so its tricky to judge the effectiveness since every modification is a possible real person.

What I wanted to show you is that you can walk through the 'latent space' of faces and generate intermediary images between any two images, thus they are not simple copies.

If you train a NN with images, wouldn't you would need permission to use the training images in the first place?

I think that that will depend on the license and the courts. A NN can be seen as a storage system - these images are retrieved from the compressed and transformed data stored in the NN; a court might rule that that is essentially the same as any other storage system. Or not.

That's a look of purely holistic vision of the world, "collection of parts" is not included.

sweet, a cursed image generator.

Reply on news.ycombinator.com