There was a paper recently where a research team trained a machine learning algorithm (a GAN they called AttnGAN) to generate pictures based on written descriptions. It’s like Visual Chatbot in reverse. When it was just trained to generate pictures of birds, it did pretty well, actually.
(Although the description didn’t specify a beak and so it just… left it out.)
But when they trained the same algorithm on a huge and highly varied dataset, it had a lot more trouble generating a picture to go with that caption. Below, I give the same caption to a version of their algorithm that has been trained to generate everything from sheep to shopping centers. Cris Valenzuela wrapped their trained model in an entertaining demo that attempts to generate a picture for any caption.
This bird is less, um, recognizable. When the GAN has to draw *anything* I ask for, there’s just too much to keep track of – the problem’s too broad, and the algorithm spreads itself too thin. It doesn’t just have trouble with birds. A GAN that’s been trained just on celebrity faces will tend to produce photorealistic portraits. But this one, however…
In fact, it does a horrifying job with humans because it can never quite seem to get the number of orifices correct.
It’s fun to ask it to draw animals though. It knows the texture of giraffes, but not quite exactly their shape. And it knows that boats are on the water, but not necessarily that they are boats.
It also (like many other image recognition algorithms) gets a bit confused about the difference between sheep and the landscapes they’re found on. Other algorithms recognize sheep in pictures of empty green fields. And this one, when asked to draw sheep…
That’s different, though, from asking it to draw *a* sheep. In that case, it knows exactly what to do. It draws the sheep, and then just to be safe it fills the entire planet with wool too.
It really likes drawing stop signs and clocks. Give it the slightest opportunity to draw one, and it will chuck those things all over the place.
Other than its horrifying humans, this algorithm can actually be pretty delightful.
I had way too much fun generating these and ended up with way more than would fit in this one blog post. I’ve compiled a few more of my favorites. Enter your email and I’ll send you them (and if you want, you can get bonus material each time I post).
once you get the hang of it (a good tip is you don’t need to write sentences, you can just list stuff) you can just use it like an instant body horror generator, watch:
Roses are red, that much is true, but violets are purple, not fucking blue.
I have been waiting for this post all my life.
They are indeed purple, But one thing you’ve missed: The concept of “purple” Didn’t always exist.
Some cultures lack names For a color, you see. Hence good old Homer And his “wine-dark sea.”
A usage so quaint, A phrasing so old, For verses of romance Is sheer fucking gold.
So roses are red. Violets once were called blue. I’m hugely pedantic But what else is new?
My friend you’re not wrong
About Homer’s wine-ey sea!
Colours are a matter
Of cultural contingency;
Words are in flux
And meanings they drift
But the word purple
You’ve given short shrift.
The concept of purple,
My friends, is old
And refers to a pigment
once precious as gold.
By crushing up molluscs
From the wine-dark sea
You make a dye:
Imperial decree
Meant that in Rome,
to wear purpura
was a privilege reserved
For only the emperor!
The word ‘purple’,
for clothes so fancy,
Entered English
By the ninth century
.
Why then are voilets
Not purple in song?
The dye from this mollusc,
known for so long
Is almost magenta;
More red than blue.
The concept of purple
is old, and yet new.
The dye is red,
So this might be true:
Roses are purple
And violets are blue
.
While this song makes me merry, Tyrian purple dyes many a hue From magenta to berry And a true purple too.
But fun as it is to watch this poetic race The answer is staring you right in the face: Roses are red and violets are blue Because nothing fucking rhymes with purple.
Hirple – To limp or walk awkwardly
Cirple – An old Scots word for the hindquarters of a horse
“i am a monument to all your sins” is such a fucking raw line for a villain it’s amazing that it came from halo, a modernish video game, and not some classical text or mythos
classic texts have nothing on the crazy people come up with in modern times tbh
“I survived because the fire inside me burned brighter than the fire around me.”
– Joshua Graham, Who Is A Fallout New Vegas NPC, Something Most People Throwing This Quote Around Don’t Realize
“If the world chooses to become my enemy, I will fight like I always have.”
– Shadow the Hedgehog in what is widely considered one of if not the single worst game in the Sonic the Hedgehog franchise
this is the source for this text and it haunts me on a regular basis
i love cutthroat kitchen but bingewatching makes it really stand out how often alton brown refers to himself as ‘daddy’ and makes contestants wear spreader bars
well i guess this is it 