Giving colors to the third reich

Giving colors to the third reich

A couple of researchers from the University of California trained a Convolutional Neural Network to predict the colors in black and white images.

research paper


How do I use this?
First of all, this neural net runs on python with opencv and imutils so you'll have to start by installing those libraries first. You can just use pip for that:

pip install opencv-python
pip install imutils
install pip if you don't have it

How do I run the model?
Just type the following in a console, inside the folder containing the .py files:

python \
–prototxt model/colorization_deploy_v2.prototxt \
–model model/colorization_release_v2.caffemodel \
–points model/pts_in_hull.npy \
–image images/your_image.jpg

After that you'll be shown with two windows, one showing the original image and the second showing the colorized image, you just have to press any key to close both windows but remember to right click on the image if you wish to save it.

WARNING: Video files can be very heavy to process
But if you want to give it a go then you should just run:
python \
–prototxt model/colorization_deploy_v2.prototxt \
–model model/colorization_release_v2.caffemodel \
–points model/pts_in_hull.npy \
–input images/your_video.webm

This one should give you a look at the colorized frames, those frames will be stored in a folder called frames and you can join them to make a video with the "" file, just remember to adjust the framerate.



Attached: original_heil01.jpeg (1024x576 719.41 KB, 782.54K)

Other urls found in this thread:

I default my phone to black and white. Less distracting. Immunity from goreposting. Post some pictures already.

Attached: french_hh.jpeg (620x388 72.64 KB, 422.33K)

Getting out my collection

Looks like it has never conceived of anything being red.

Attached: Colorized_h_0_01.png (968x681 870.48 KB, 83.22K)

My blood for the flag. The 1000 year Reich will forever be when media dies. Ask anyone who watches judenpresse week in week out.

Attached: Francis.png (385x385, 546.83K)

They forget by the next week. What a state of art of.

Attached: Colorized_f_tr.png (770x427 486.35 KB, 52.49K)

Attached: Colorized_easternfront.png (800x422 401.01 KB, 15.48K)

Have a bump, user, because this is fucking rad.

Attached: ebin.png (232x266, 3.91K)

Attached: Colorized_arch.png (763x507 495.83 KB, 498.07K)

Attached: Colorized_dinner.png (722x406 346.63 KB, 47.88K)

Attached: Colorized_news.png (620x330 243.72 KB, 28.23K)

Welp it's getting pretty late for me, I'll continue tomorrow and also I'll figure out a way to get past 8ch's retarded file uploader to show a video I processed.

Attached: Colorized_modelcar.png (1366x765 1.18 MB, 135.16K)

What color does it give for Hitler's eyes?

the deepest of blues

Attached: output_test.mp4 (500x280, 631.5K)

Attached: Colorized_bb.png (771x998 1.18 MB, 435.08K)

Attached: Colorized_xmass.png (902x661 775.46 KB, 297.37K)

Attached: Colorized_flowers.png (658x642 553.01 KB, 121.81K)


Happy 4/20.

I like how the last one turned out.

Attached: Colorized_dg.png (1201x1192 1.44 MB, 224.23K)

I like the one with the christmas tree. Germany seems so lively back then.

This is like a German only Christmas.

I'll go looking for more pictures.

Attached: Colorized_sub.png (318x275 198.13 KB, 66.38K)

This is one is for you Alfred.

Attached: Colorized_rosenberg.png (566x566 264.08 KB, 37.02K)

Attached: Colorized_bb3.png (1002x657 1.1 MB, 366.81K)

Attached: Colorized_himm.png (472x643 602.96 KB, 109.72K)

Attached: Colorized_beauty.png (928x1122 1.02 MB, 168.68K)


Attached: Colorized_kid02.png (1810x961 1.93 MB, 321.88K)

Attached: Colorized_talk.png (159x325 101.92 KB, 18.85K)

Attached: Colorized_deer.png (590x782 502.41 KB, 111.18K)

Is you mouse now sticky?

Attached: Colorized_bffs.png (962x641 1.12 MB, 198.64K)

Some of these results are obviously shitty. You could crop parts that didn't work properly and try that piece seperately.

I already tried that with many images but the result was either the same or even worse, the thing is, this neural net needs a lot more training samples and time, I'm pretty much testing it on every image I can find and seeing what happens.

Attached: Colorized_prd.png (831x605 587.03 KB, 82.54K)

Can we retrain the model? Use the one they have as a base but feed it a bunch of moden re-enactment photos.

Yes, you can do that, all you gotta do is call all the components of the model that were saved in the last checkpoint (the files in the model folder) and run a training session with new images to feed the network. The only problem is that I don't know yet how to do that with caffe models, gonna have to study that.

Does it work like deepfakes? Used to have that pumping for a while.

Close to it, this model uses a simple Encoder -> Decoder architecture, which honestly doesn't give too good results, however it is more faster and less expensive than what deepfakes use.

Deepfakes use GANs (Generative Adversarial Networks) to create their output, in short, it consist of one network creating fake data and another network inspecting said data and determining if it's real or fake, you basically have a network training hard to jew the other one to output a fake image that looks legit to us. This the kind of network I'd recommend for the colorization problem, it would give better results but as you've might imagined it requires a ton of GPU power and millions of samples.

This is really cool OP, thanks for sharing. It seems blurry spots and skies don't always turn out well but overall it is very impressive technology. Posting some of the better ones.

Attached: 3.PNG (497x614 688.56 KB, 154.58K)

Thank you too, those are nice results. The model also has trouble whenever there are too many faces together. But it somehow recognizes trees and leaves quite easily.

Alright these will be the last ones from me, I gotta go, whoever wants to give it a try just toy around with the files I shared and post your best results.

Happy 4/20

Attached: Colorized_rom4.png (800x415 470.16 KB, 54.76K)

Hitler was a homosexual.

4/20 is for getting stone as fuck, not worshipping some dead gay Nazis.

Attached: 4C9AE16E-B733-44EA-8ADD-A318746DE645.jpeg (1536x2048, 373.63K)

using script:

@echo off
set /P id=Enter image name:
@echo off
-prototxt model/colorization_deploy_v2.prototxt
-model model/colorization_release_v2.caffemodel
-points model/pts_in_hull.npy
-image images/%id%
pause null

to launch it, getting
the following arguments are required: -c/–points
error, any help?

oh look boys, i found another kike.
Heydrich! Get der flammenwerfer! Deiser moment
burn untermensch

Attached: Heydrich.get the flamenwerfer.mp4 (1280x720, 369.64K)

Try with this format

python \
–prototxt model/colorization_deploy_v2.prototxt \
–model model/colorization_release_v2.caffemodel \
–points model/pts_in_hull.npy \
–image images/%id%

Just a reminder, each argument must be called with double '–'

That one came out nicely. They look ever more patriotically stimulating in colour.

I've been fucking around for 45 minutes or so and can't figure this python shit out. sorry lads, can't contribute due to retardation. thanks for the great work.

Attached: Untitled-2.png (150x198, 54.56K)

Ty dudes, fixed it

Attached: Colized caped.PNG (721x621, 868.6K)

Make a .bat file with the script:

@echo off
set /P id=Enter image name:
python –prototxt model/colorization_deploy_v2.prototxt –model model/colorization_release_v2.caffemodel –image images/%id% –points model/pts_in_hull.npy
pause null

in the folder containing the file and run the batch file.
Then run the batch file and type in the name* of the image you want to convert.
*include the .jpg, .png, etc at the end and make sure the image is in the images folder.

Attached: furhrer colorized.PNG (570x802, 1021.58K)

>File "", line 5, in
Thanks anyway user but i'm still too retarded. I've gone back and looked through the command and everything and don't know what the fuck this "numph" thing is. I found a site that does it pretty good online though. see pic related.

Attached: colorized-image(2).jpg (400x533, 24.65K)

pip install numpy

Alright boys, with my minimal python knowledge;
I've figured out how to make this this about 10
times more convenient.


1) Create a folder named "color_images" in the main directory
2) Go into the "" and paste the lines:

sep_img_path = args["image"].split("/");
sep_img_name = sep_img_path[1].split(".");
img_name = sep_img_name[0];
print("name parsed: "+img_name)
cv2.imwrite("color_images/color_"+img_name+".png", colorized)

on the second to last line of the file.
3) Use the script I previously posed and watch as all of your
colorized images are saved in the color_images folder without the snipping tool.

Have Fun & Heil Hitler

Didn't work bro. I didn't have PIP so i got a file online "" and installed that with python. it came up that it had been added and everything but i still get the same error when running your script from earlier about "numpy". not sure what i'm going wrong.

What version of python are you using?
The most stable from what I remember is currently 3.5.2

I will try out with 3.5.2 but then I'm giving up (kek)
appreciate your help though user and not laughing at me and calling me a faggot for not understanding

These are brilliant, keep them coming

No problem dude, hope it works

Attached: color_h6.png (1827x2699, 8.12M)

Some turn out better than others

Attached: color_h10.png (500x925 993.8 KB, 650.66K)

yeah, i'm going to give up mate. i think it's got something to do with this pip shit. i can't be fucked anyhow, but do appreciate your help though.

For any other retard anons, try below

Attached: colorized-image.jpg (1920x1080 253.13 KB, 252.54K)

its not in your PATH system variable. you'll need to add the pip executable's path to your system's path so your cmd knows where to look for it. quick google search for "add python to system path" will work for you.

Attached: ClipboardImage.png (1034x588, 41.42K)

nah mate, i changed the directory manually when i was seeing if i could get it to work or not. thanks for that though, i will keep it in mind for other useful cases.

Here is a link to the modded on if you didn't do it already.

hey, what do you know, the retard user finally figured it out. Cheers for all your help mate.

Attached: color_anne.png (750x951 832.04 KB, 199.21K)

I'm sorry, user, but if you can't figure out the plebby python then you're not an aryan

it's figured out and yes, 100% nordid phenotype

Attached: color_2.png (500x690 695.11 KB, 138.02K)

Another one dodges the oven

What's the problem? You throw a lasso around the not colored part you want colored and select to colorize. Pick your color. I'm talking in Gimp. Maybe Photoshop has similar methods.

Attached: 8.PNG (939x619 1.09 MB, 80.34K)

For example. You didn't get Hitler's sleeve. So I lassoed his sleeve and colorized it. Hue 243 Saturation 9 Lightness -48. You do this for each individual color. It's time consuming of course. Not automated.

Attached: 7ef8d1b5e340f90d3ddf9b6cb923308f950fd8aae53089fd19c2935770840ea2 (1).png (500x690, 558.82K)

At least the lasso algo is automatic, somewhat. That's what makes it possible, really, though someone probably did one pixel by pixel with windows 3.1 or the like.

Beautiful thread, good sir

Attached: 772dc9aa50b1a65235b597768fe8fbfcede85f75132a422d2e333fa92f6a6ff5.png (589x459, 365.89K)

Despite being a chemist, I'm a tech retard. Can someone please do some Evolas for me? Thanks

Attached: index.jpeg (500x500 90.47 KB, 43.05K)

No shit bro. I tried working on a image yesterday and was doing it color by color. It's going to take time if you want it done properly (pic related was a half hour or so). This algorithm is certainly not perfect; it does make the images look better though. It's not supposed to be perfect, yet, until the AI gets better anyway. For now though, this gets the job done quick and makes a vast improvement on the initial b&w image.

Attached: before and after.png (737x533, 318.44K)

Attached: color_e.png (500x500 1.04 MB, 384.27K)

MY NIGGER. Danke and happy 4/20!!

How do you guys colorize it? Through photoshop?

You know that picture is a 'cosplay', right? It's from a year or two ago.

yeah, i can't find it in colour though. you got it faggot?

"Who the fuck are you"
"I'm the guy that's here to save the goddamn day"

Attached: cage_bliss.gif (403x219, 462.99K)

MS paint

Bump, I'll try later


amazing thread

>i'm a drawfag
what does this mean, user?


Images in black and white usually feel 'disconnected' from people living today in a sense. If we took images of Hitler's Germany and gave them color it would actually go far in humanizing a nation and a man who has been dehumanized by history.

Appreciate the pics very much.

This is great

This is true, and I initially assumed it was the point of this thread.

Attached: color_2.png (1200x720 1.26 MB, 146.71K)

I get "is not recognized as an internal or external command" etc. with:


Show the exact command you run
Show contents of pwd
Which OS are you using?
If you pasted from here make sure to replace the weird unicode dash with two dashes

Attached: Colorized_screenshot_27.04.2019.png (558x768, 569.41K)

python \
–prototxt model/colorization_deploy_v2.prototxt \
–model model/colorization_release_v2.caffemodel \
–points model/pts_in_hull.npy \
–image images/young.jpg

I don't know what pwd is.

Windows 10