Mario Klingemann💧💦 Profile Banner
Mario Klingemann💧💦 Profile
Mario Klingemann💧💦

@quasimondo

58,212
Followers
2,225
Following
6,195
Media
40,272
Statuses

Artist, Neurographer, AI Prompteur, Purveyor of Systems, Data Dumpster Diver, Information Recycler

Munich, Germany
Joined April 2007
Don't wanna be here? Send us removal request.
Pinned Tweet
@quasimondo
Mario Klingemann💧💦
8 years
Peter sees the computer. "But the machine only creates what humans have taught it to," says Peter. "So do you," says Mummy.
Tweet media one
104
1K
4K
@quasimondo
Mario Klingemann💧💦
5 years
Saw a pretty moiré effect in the Manchester airport smoking area and tried to replicate it in a shader. #CreativeCommute
13
374
2K
@quasimondo
Mario Klingemann💧💦
6 years
Some real time face play, checking if the models have learned something new.
35
527
2K
@quasimondo
Mario Klingemann💧💦
6 years
Fingerplay (Take 1)
20
489
2K
@quasimondo
Mario Klingemann💧💦
6 years
Turn your face into an oil painting in real-time* * ᵃᵖᵖˡᶦᶜᵃⁿᵗˢ ᵐᵘˢᵗ ᵇᵉ ʷʰᶦᵗᵉ ᵐᵃˡᵉ ᵃⁿᵈ ᵒᵛᵉʳ ⁴⁰ ʸᵉᵃʳˢ ᵒˡᵈ
29
510
2K
@quasimondo
Mario Klingemann💧💦
5 years
I didn't know who Harry Styles was until today when I learned that he's the guy who blocked off the entire St. Abbs harbour and prevented us to enjoy our fresh crab rolls. 😫🚧🕺🦀🥖
139
202
2K
@quasimondo
Mario Klingemann💧💦
7 years
Facefeedback II
18
369
1K
@quasimondo
Mario Klingemann💧💦
8 months
Computerworld, 1968
Tweet media one
30
333
1K
@quasimondo
Mario Klingemann💧💦
3 years
Maybe try to appreciate just the textures without thinking too much about what you see. #CLIP #VQGAN
Tweet media one
56
98
1K
@quasimondo
Mario Klingemann💧💦
4 years
Current progress on mapping music to facial expression vectors. #StyleGAN2 #realtime Song: "Triggernometry" by Kraftamt, 2014
71
354
1K
@quasimondo
Mario Klingemann💧💦
6 years
With my similar scene finder I can quickly find interesting training material to fine tune my Old Masters face model and it adapts pretty fast.
28
249
1K
@quasimondo
Mario Klingemann💧💦
6 years
I think I told you that I really suck at drawing.
29
330
1K
@quasimondo
Mario Klingemann💧💦
4 years
Another attempt at a longer piece. An imaginary Jerome K. Jerome writes about Twitter. All I seeded was the title, the author's name and the first "It", the rest is done by #gpt3 Here is the full-length version as a PDF:
Tweet media one
38
373
1K
@quasimondo
Mario Klingemann💧💦
2 years
Somebody should have told me in 1984 that this was generative art. But they called it "Games with Computer Graphics".
36
103
1K
@quasimondo
Mario Klingemann💧💦
5 years
I should save these for a rainy day, but I can't help it, they want out now. #BigGAN #Teddy
27
221
1K
@quasimondo
Mario Klingemann💧💦
4 years
I started collecting some direction vectors in #StyleGAN2 that try to change only a certain property of a generated face. But it's hard to get clean ones. Here's head rotation:
28
174
1K
@quasimondo
Mario Klingemann💧💦
6 years
So spectral normalization is a mixed bag. Overall it seems to make everything more consistent and there are no blatant artifacts anymore. But it couldn't fully heal the one that had settled in the top-left corner.
15
257
1K
@quasimondo
Mario Klingemann💧💦
7 years
Here's what happens if you badly train a neural network to predict the next frame in a video and give it a still image to start with:
29
433
979
@quasimondo
Mario Klingemann💧💦
5 years
Optimizing for unrecognizability #BigGAN
Tweet media one
Tweet media two
Tweet media three
Tweet media four
27
225
978
@quasimondo
Mario Klingemann💧💦
7 years
This might be one the creepiest clips that I have made so far. Yes Sir, that's my baby. #FaceFeedback
28
295
946
@quasimondo
Mario Klingemann💧💦
5 years
I am proud that I had the chance to work with #MassiveAttack creating neural visuals for their new #Mezzanine anniversary tour which started yesterday. (I ripped that video from in_cmmn on instagram).
36
98
987
@quasimondo
Mario Klingemann💧💦
5 years
Automatically generated #BigGAN music video with "EMT" by @joshmobleymusic
22
219
840
@quasimondo
Mario Klingemann💧💦
6 years
Another pretty sketchy clip of the Old Masters model in action.
16
266
862
@quasimondo
Mario Klingemann💧💦
4 years
Some pose-to-pose puppeteering with #StyleGAN2 . Just the head so far, but obviously expression mapping is the actual target.
20
134
843
@quasimondo
Mario Klingemann💧💦
3 years
The Hindenbird Disaster #CLIP #VQGAN
Tweet media one
25
66
819
@quasimondo
Mario Klingemann💧💦
7 years
"Generative Computergraphik" from 1969 by Georg Nees arrived in my mailbox today. Here's some pioneering generative art written in ALGOL.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
10
223
806
@quasimondo
Mario Klingemann💧💦
3 years
I do not really care much for "pretty" generative art. I want something that has an interesting concept, is algorithmically challenging and ideally so complex that I cannot reverse engineer its mechanism right away. Unfortunately I don't see a lot of that these days.
136
44
781
@quasimondo
Mario Klingemann💧💦
7 years
Emojis at 10000x magnification.
Tweet media one
Tweet media two
Tweet media three
10
199
746
@quasimondo
Mario Klingemann💧💦
7 years
Generating faces from a sketch. I trained a pix2pix net on 1500 #bldigital faces. Left input, right output.
14
284
730
@quasimondo
Mario Klingemann💧💦
3 years
This prompt is a keeper. #CLIP #VQGAN
Tweet media one
37
53
709
@quasimondo
Mario Klingemann💧💦
5 years
Now that I have my local version of #BigGAN running, exploring the wilderness and weirdness of its latent space will become a bit faster and comfortable.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
20
154
673
@quasimondo
Mario Klingemann💧💦
2 years
If you just woke up to and wonder what happened to the hic et nunc website and its "discontinued" twitter account here's some info what is going on as far as I know:
35
238
642
@quasimondo
Mario Klingemann💧💦
5 years
Found a hideaway in the wastelands of #BigGAN .
Tweet media one
Tweet media two
20
120
636
@quasimondo
Mario Klingemann💧💦
5 years
Inverse #BigGAN search for a previously generated image across the entire 1128-dimensional latent space starting from a random vector.
16
145
627
@quasimondo
Mario Klingemann💧💦
4 years
Some recreational coding on the train to Berlin.
8
28
622
@quasimondo
Mario Klingemann💧💦
7 years
The model has evolved over night.
25
168
592
@quasimondo
Mario Klingemann💧💦
5 years
Latent Lisa, found in #StyleGAN #FindingMona
15
149
614
@quasimondo
Mario Klingemann💧💦
6 years
I'm training a next-frame-prediction #pix2pix model on some fireworks footage.
6
103
597
@quasimondo
Mario Klingemann💧💦
5 years
A Collection of Broken Records. #BigGAN
17
103
584
@quasimondo
Mario Klingemann💧💦
4 years
Hey #gpt3 , the sentence: "A giraffe in a forest with trees in the background" is correct but boring. Make it sound more interesting.
Tweet media one
Tweet media two
16
128
579
@quasimondo
Mario Klingemann💧💦
6 years
Well, you know the drill...
14
147
567
@quasimondo
Mario Klingemann💧💦
6 years
Finding movie scenes by example. After recording a scene with my camera the machine analyzes it and returns the visually most similar ones.
10
109
567
@quasimondo
Mario Klingemann💧💦
7 years
Here is my nasty visage run through a chain of two GANs, the first one "normalizes" it, the second one re-imagines the lost details.
16
169
562
@quasimondo
Mario Klingemann💧💦
5 years
I've created an experimental GAN architecture I call #RecuResGAN or "Recursive-Residual GAN" and I am pretty astonished that: - it works at all - how well it works across a pretty wide range of scales. - it is just 15% the size of a comparable #pix2pixHD model
Tweet media one
Tweet media two
Tweet media three
Tweet media four
17
79
561
@quasimondo
Mario Klingemann💧💦
5 years
The journey is the trip. Or something like that.
14
139
547
@quasimondo
Mario Klingemann💧💦
7 years
This CGAN neural network I trained overnight has its own ideas what "Enhance!" means. Left: input, right: output
Tweet media one
Tweet media two
Tweet media three
Tweet media four
28
173
535
@quasimondo
Mario Klingemann💧💦
3 years
Probably a bit of an overkill just to scan slides. #robotporn
34
35
518
@quasimondo
Mario Klingemann💧💦
4 years
The autobahn is open. The pioneer silently weeps and reminisces how there used to be jungle here just 3 years ago.
@Adobe
Adobe
4 years
Neural Filters is a major breakthrough in AI-powered creativity and the beginning of a complete reimagination of filters and image manipulation inside @Photoshop 🤩: #AdobeMAX
188
2K
9K
17
73
509
@quasimondo
Mario Klingemann💧💦
2 years
@HerbertWFranke Great to have you here! Your fanclub.
Tweet media one
4
7
501
@quasimondo
Mario Klingemann💧💦
6 years
I have spent too much time in front of my Uncanny Mirror - it has learned my glasses, my shirt and my mobile phone. @mediacityseoul
9
82
477
@quasimondo
Mario Klingemann💧💦
6 years
The magic of LCD handheld screen layouts.
Tweet media one
Tweet media two
Tweet media three
17
156
476
@quasimondo
Mario Klingemann💧💦
7 years
Our customer support team is ready to help.
Tweet media one
Tweet media two
12
56
455
@quasimondo
Mario Klingemann💧💦
7 years
This is a fully generated clip, made from the random interpolation, transformation and scaling of the original Françoise Hardy face markers.
11
180
473
@quasimondo
Mario Klingemann💧💦
4 years
Now would be a great time to try out Universal Basic Income for a year.
8
101
472
@quasimondo
Mario Klingemann💧💦
3 years
We made our first million at @hicetnunc2000 .
@ObjktsX
〇BJKTsX{yz}
3 years
H=N silently passing through 1M XTZ mark in artwork trades since its launch. @hicetnunc2000 👉 :
Tweet media one
16
75
367
36
71
472
@quasimondo
Mario Klingemann💧💦
6 years
Another round of fine-tuning, starting from a state where the artifacts had not shown yet. This model uses batch norm in the deep residual layers, instance norm in the surrounding ones and spectral norm in the discriminator.
10
94
442
@quasimondo
Mario Klingemann💧💦
5 years
"Art for a Proctologist's Waiting Room" #10yearsago #BlastFromThePast
Tweet media one
Tweet media two
9
46
439
@quasimondo
Mario Klingemann💧💦
6 years
This might show it a bit better. Left is input, right is output. My latest progress in neural enhancement using a modified #pix2pix architecture which tries to achieve better details. The training data right now is female faces only.
Tweet media one
Tweet media two
14
110
433
@quasimondo
Mario Klingemann💧💦
4 years
Idle Loop #StyleGAN2
17
69
434
@quasimondo
Mario Klingemann💧💦
5 years
Realtime three GANs plus camera feedback loop. #InterstitialSpace
4
96
431
@quasimondo
Mario Klingemann💧💦
3 years
No, "Famous artist finding a new way to make more money in an already booming market" is not a game changer. "Artists from all backgrounds creating a thriving decentralized ecosystem and community from scratch on the blockchain without having any funding" is one. @hicetnunc2000
20
82
421
@quasimondo
Mario Klingemann💧💦
5 years
Josh Mobley @joshmobleymusic just sent me a new song called "A Walk in the Park" and I had my algorithm give it the #BigGAN treatment. Since it can create an infinite amount of versions it's hard to pick one, but I really like what it does at 1:08.
21
131
427
@quasimondo
Mario Klingemann💧💦
4 years
Ungrateful lazy bastards. Just a few years ago you could be happy if there was any code accompanying a paper.
Tweet media one
24
29
424
@quasimondo
Mario Klingemann💧💦
2 years
It's no secret that Art Basel Miami is all about NFTs this year. Make sure to check in at the @tezos booth and get your surreal self-portrait NFT minted by my installation: @TheArtNewspaper
Tweet media one
12
96
418
@quasimondo
Mario Klingemann💧💦
4 years
Congratulations, your lockdown has been extended by 4 weeks. #StyleGAN2
18
92
415
@quasimondo
Mario Klingemann💧💦
5 years
I wish this glitch version of my work were intentional, but it's an actual glitch.
Tweet media one
Tweet media two
17
22
413
@quasimondo
Mario Klingemann💧💦
5 years
This actually does stand out amongst the current avalanche of "I trained a GAN on everything and the kitchen sink". Nice one!
@maxbraun
Max Braun
5 years
What happens if you train a GAN on the @eBoyArts database? Infinite pixel nightmares, that's what.
64
2K
7K
3
40
417
@quasimondo
Mario Klingemann💧💦
5 years
Oh, this is working better than I had hoped for. A different type of #BigGAN evolutionary search. #neurography #neurealism
Tweet media one
Tweet media two
Tweet media three
Tweet media four
12
81
405
@quasimondo
Mario Klingemann💧💦
6 years
OMFG!!!!! @lumenprize
Tweet media one
59
12
408
@quasimondo
Mario Klingemann💧💦
2 years
After a few hours of figuring out how to translate bezier curves into robot moves that don't wiggle, stutter or run into singularities I finally managed to have it sketch a #CLIP generated "Mona Lisa".
20
38
407
@quasimondo
Mario Klingemann💧💦
3 years
Current state of my pair-to-pair model. 🤔
14
34
402
@quasimondo
Mario Klingemann💧💦
5 years
Sketch for a new work - "Three Latent Body Problem"
14
54
401
@quasimondo
Mario Klingemann💧💦
4 years
I had to try my own method for this problem. Not sure if you can call it an improvement, but by simply starting the gradient descent from different random locations in latent space you can already get more variation in the results.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
11
70
401
@quasimondo
Mario Klingemann💧💦
6 years
This model tries to learn fluid dynamics by watching a video and then generating it via next-frame prediction. #pix2pixHD #feedbackloop #GAN
13
70
393
@quasimondo
Mario Klingemann💧💦
2 years
Some #genart history: Algorithm 133 by Peter G. Behrenz from 1962 might be the first documented random number generator used in generative art. It can be found in Georg Nees' book "Generative Computergraphik"
Tweet media one
Tweet media two
15
62
390
@quasimondo
Mario Klingemann💧💦
7 years
Here's how they are made: The input is a 32-dimensional human pose vector as visualized in the upper corner, the output is what you see.
Tweet media one
Tweet media two
Tweet media three
15
137
385
@quasimondo
Mario Klingemann💧💦
4 years
I grew up here.
Tweet media one
28
27
386
@quasimondo
Mario Klingemann💧💦
7 years
Tweet media one
6
62
389
@quasimondo
Mario Klingemann💧💦
2 years
So many variables in just drawing a straight line with ink + brush (+ robot). 🤯
12
22
394
@quasimondo
Mario Klingemann💧💦
3 years
WARNING: Unfortunately somebody has figured out a new nasty copyminting scam on @hicetnunc2000 which is very deceptive since it shows the real artists as the creator, but they did not actually mint it. Do not buy from this wallet: tz1ch5WHJDZbbFR8R98cDkriVMwFP6RN9it2
38
207
387
@quasimondo
Mario Klingemann💧💦
6 years
A restoration attempt.
Tweet media one
9
49
393
@quasimondo
Mario Klingemann💧💦
4 years
Ah, the cat is out of the bag. 😊
@MassiveAttackUK
Massive Attack
4 years
6
149
673
28
29
382
@quasimondo
Mario Klingemann💧💦
1 year
Cloning and installing #ChatGPT inside of itself and then running the experimental concious_chat.py script. 😂
Tweet media one
Tweet media two
Tweet media three
Tweet media four
25
48
378
@quasimondo
Mario Klingemann💧💦
4 years
My new model makes it really hard to decide what to throw away, when the worst results I get look like this. #aiart #neurography
Tweet media one
Tweet media two
Tweet media three
Tweet media four
21
34
377
@quasimondo
Mario Klingemann💧💦
2 years
"Plastic Trash in a Storm at Sea"
Tweet media one
15
49
374
@quasimondo
Mario Klingemann💧💦
2 years
Various #dalle explorations.
Tweet media one
Tweet media two
6
23
367
@quasimondo
Mario Klingemann💧💦
6 years
More fun with uv map extraction.
Tweet media one
Tweet media two
10
42
356
@quasimondo
Mario Klingemann💧💦
5 years
This was the inspiration.
Tweet media one
3
17
363
@quasimondo
Mario Klingemann💧💦
2 years
A few motion studies with quick brown foxes.
Tweet media one
12
27
354
@quasimondo
Mario Klingemann💧💦
7 years
Q: "Where do the little faces come from?" A: "Well, when two not very well trained GANs meet and enter a feedback loop it just happens"
9
108
347
@quasimondo
Mario Klingemann💧💦
6 years
If all you have is an automatically generated uv map everything looks op-artable.
8
53
340
@quasimondo
Mario Klingemann💧💦
5 years
A friendly reminder to everyone having fun with Snapchat these days, where you saw it first (including the popular face palming).
@quasimondo
Mario Klingemann💧💦
6 years
With my similar scene finder I can quickly find interesting training material to fine tune my Old Masters face model and it adapts pretty fast.
28
249
1K
12
45
342
@quasimondo
Mario Klingemann💧💦
4 years
I'm having too much fun with the #U2net + #StyleGAN2 combo. So one more in which the ears fall off. Song by Kraftamt: "Liberation" (2020)
11
47
342