Jewdar (from the dead OP)

I do machine learning for fun and profit, and I have some spare cash to train a tensorflow network on the GCP machine learning API, but not an overwhelming amount. I mostly do predictive modeling, so computer vision is not necessarily my forte, but I'm more than qualified to work on it. If anons collect more corpus data, I would be willing to help build this.

Also I have a hypothesis that there is enough data in jewish speech patterns to detect classify them via audio. If I had 10,000 hours of jew and non-jew audio clips I think it would allow for a pretty accurate model using a recurrent neural network.

...

checked, the level of tech available to the average user now presents itself as a challenger to (((their))) lvl of control mechanisms and smoke screens. the golem they built (high level tech) is biting them in the ass

...

doesn't every jewish story about golems end with the golem rising up against its masters and destroying them?

Not that it is a bad idea, the only problem is that you would have to do every day a different pattern, before the machine learns how to beat this shit.

They stole.

Just think for a second, are the indigenous populations the end, or the beginning of the civilization?

A system that exists in an atemporal scene, with isolated groups that never meet nothing that challange their perceptions, no technology, no nothing, and yet, if they do their corpse paint in the right way, they can shoot everybody in hong kong without even being identified by the local police.

maybe war paint is a cultural relic from an ancient society like what the jews want to turn ours into, and long after it was overthrown and crumbled and people returned to a primitive way of life they kept painting their faces the way their ancestors did before going into battle, not knowing that their ancestors only did it to confuse the AI attack drones defending the ancient power elite.

can't hide the nose from 3d imaging

So we're going to be like that shitty game Omikron?

>mega.nz/#!MJ8kmbYD!T7ahVYln8ysnA5EoXULug5ODKz_DnYCOWUuTCrdtiqE
i will risk downloading a virus because i hate the jew

The Jews have already turned us into what they want. This present “civilization” has been so degenerated that it would be a blessing to go back to throwing spairs and using war paint.

Use DLIB to center, scale and rotate your faces. (224x224). Push results into positive and negative directories. Find Keras VGG Face on GitHub. Use ImageDataGenerator to load the images and don't do a bunch of weird augmentation shit. Extract features from the images using VGG face and save to a numpy array. That will be the input array for the model you will train. I used 4 full connected layers with 25% dropout in between (512 > 128 > 32 > 1, I think is what I used). Expect your model to not generalize nearly as well as validation accuracy unless you have excellent data. Your training data is everything, and you need a lot of images (probably 1K at least. 10K is better) Really, most of the work is getting images that don't suck.

Seriously, this will work. Maybe not perfectly, but probably better than a person and definitely better than random. It's just applied stats and linear algebra. I'll be able to post some very similar work around the 18th. I'll drop some source for a handy scraper for Google image searches maybe mid next week. If you can find source for a racial classifier before then, just use that and feed it Jews vs Europeans. I don't know how you're going to get a large number of random probably-not-Jewish pictures. Maybe FOIA DL pics from the DMV of a rural county?

Keep it stickied. This is our magnum opus. The Halls of CNN will be filled with terrified whispers of the name user.

How to tell difference in Jewish names from European names? For example, is Schuessler different from Schussler? -Bergs compared to a European type -berg like Forsberg? Also have a name chart I saved.

< leave set on 1.2 ALL THE TIME

Too late?

Just shit camouflage.

Is this a fucking joke? We have memos being released that are exponentially way bigger then Watergate.

This is bullshit for being stickied.

Looks like someone needs a ban.

...

Why Keras and not Convolutional?

Yep. Camouflage is interesting, and surprisingly expensive… pic related - cheaper stuff from them.

flecktarn is very affordable and effective depending on where you are

Anything helps.

We still need data for the images, and as the old thread claimed, the datasets are filled with false positives and negatives.
The DLib is for facial landmarks right?
Also VGGFace? Are there others like OpenFace and Deepface that could work better?

Keras is the library, like Tensorflow and Theano. All three of these allows for multi-layer perception, Convolutional NN and LSTM/GRU styled neural networks.

Really makes you think.

Would be banned as it counters surveillance cameras.

Nice thread, should be stickied.
Someone implement this into a browser extension so it can filter all of them with ((())).

Absolutely what this user says, he knows precisely what he's talking about.
Luckily we have the power of collective autism on our side. I am setting up something like the google captcha so that anons can select if a scraped picture is a human/jew/other, and therefore build an accurate corpus.

Would make a hilarious replacement for the Holla Forums captcha system once the net is trained good enough.

Standard gradient descent/linear regression algorithms could be relatively easily combined with qualitive data if anons could start compiling things. Surnames, typical professions, you name it. All this can be combined pretty easily to get a far more mature AI framework if people really got to work on the qualititative side of things.

too late - you missed first post

Kikery at work here lads
Sure thing, Schlomo!

so lets get at it
let me start by checking those dubs

Here's a thought: there are jews that write text in a very specific style. I'll just call it 'pilpul stile'. Just reading it makes my blood boil. You could analyze texts written by jews with rabbinical-appearance. Typical deconstructive hair-splitting crap in a stream of loosely-tied jumps. Then the computer can identify them. At least procuring and formatting the necessary data is a lot easier.

It sounds like it's time to rebel then, you cuck.

I was wondering what happened to this. I remember it had 88% success rate according to the OP. Then it vanished.

Hook this up to the DMV database and make DOTR even more efficient.

I don't know if there's much of a difference in the case of German surnames. For example you had Alfred Rosenberg in the NSDAP, yet in the US that surname probably belongs to a kike, case in point (((Yair Rosenberg))), some faggot who calls himself "the troller of nazis" and who, appropriately for this thread, tried to get others to create a shitty system to identify far-right trolls from twitter who posed as kvetching kikes.

They sell that stuff that looks like someone dripped paint on it at Walmart.

I'd recommend CADPAT (illegal to use in leafland unless you're a soldier though).

< not using OnionS'hare

your kidding…. colored clothes of a certain pattern are illegal?

I can't find any evidence of the claim.

This user seems fairly blackpilly, since even if it was functionally useless, it would still be good to be in the news and make good meme material even for normies.
This is the only valid part of the post. Behavioural phenotype is what's most important and what people should be educated on.

Kind of funny how shit is melting down but complete bullshit worthless shit threads like this get stickied.

Anybody notice this.

Leaf here CADPAT is sold at every army supply and its not illegal
fag

So we should be impotently gossiping about what's on the infotainment channel? What do you want me to do, specifically?

Open face might be better. I don't know from experience. I used VGG face for a racial classifier and for a criminal phrenology project that I'm still developing. Both are working so far. I'm stuck phone posting until next week, so sorry for the low effort response.

Please let this happen.

I remember the thread, tried to search for it on the archive, I also remember it disappearing after few hrs, I regreted not downloading the files.
Thank you OP, I will see what can I do about this

If you put all the jews you can find, not just celebrity kikes with rinoplasties but all the kikes you can find like at the nearest jewish temple, this thing will be as accurate as a genetic test.

You must be new here, from 4chan or something. Check the age of threads page 15+
We have over 20 pages worth of content on a single board. This allows both current news and some archive age threads to be browsed seamlessly.

Actually much harder to teach, but much more reliable once taught

t. Programmer

faceplusplus.com/face-searching/#demo

Add your jewish and nonjewish photos to the collection (maybe we could add a small note on each photo that says jew/goyim) and upon uploading a photo it will match with the top 5 most similar faces from your collection, giving you a very good idea of who is jew and who is goyim.

i tried uploading characters from Harry potter for goyim, and my childhood was ruined when it turned out the face recognizer worked, and daniel radcliffe is a jew

[email protected]
godwithus

for a collection of images of jews

There's no such thing as what technology "can't do", only what "it still doesn't do".

3D imaging is used in phones now. There's no doubt it could become cheaper to the point it's a feature of security cameras, much like night vision. I'm aware of how simple night vision is compared to 3D imaging

Think.

Don't take this image too literally. Half of the names on there are common Germanic names that can belong to goys too, because they derived their names from the same methods as described in the picture when surnames became mandatory.

Not long ago, playing a game of chess would be called "thinking". Or, consider Amazon's Orwellian grocery store they're test driving where robots look at what you pick or put back in the shelf. Or driving!

Not saying a computer can think like you and I do, but it certainly can think more than the average normalfag. It can "think" enough to detect kikes and send drones, most likely.

The word think implies the word ‘reason’. Machines still can’t reason, and never will be able to. Nor will they ever be able to express emotion. Machines are not and will never be people and will never have rights. Keep your ✡transhumanist✡ shit to yourself.

That's just semantics, isn't it? "Reasoning" can be reduced on a local level to a series of conclusions derived from base premises. Why couldn't it be so in a bigger scale? Maybe your whole life is a complex series of conclusions. I don't know that, but I do know they can perform more reasoning than you neighboor, and that should be a cause for concern, and an opportunity to harness.

Again, imagine automated drones self-replicating in an unstoppable automated holocaust. Now imagine the same happening to whites. Both things are very important and they rely on what, to many in the past, would be called "reasoning".

Maybe not, but some people think Tay expressed emotion. I don't

I hate that bunch as much as you.

All you've proven with your post is that you do not know how neural networks work or what has already been achieved. Go back to reading TV Guide grandpa.

I killed my mentor
but I did not murder him

I am almost afraid to do a 23&Me. I still have blonde hair, blue eyes, and dont have a Jew nose. But there are days I wonder what I'd do if I actually had jewish blood in me

CultState here. My Twitter (@CultStateDotCom) will tweet this post number as a base64 encode as proof of identity.

Thank you, mods, for giving this a sticky. It is essential to the progression of the Butterfly War. The original Jewdar post is here (archive.fo/XNuuN)

California Senate races are heating up and I've been very busy trying to get Google and Facebook employees to spill the beans about Silicon Valley's usage of cyber phrenology. You won't believe the influential circles I've recently been able to be a part of and I'll be seeding the chans with screenshots of my most recent infiltrations over time as proof of this claim.

The efforts detailed here (>>11241329) should be expanded upon and I'll see if I can steer a few experts in to help.

The goal isn't to make something that is perfect. The goal is to show that the intentions and the idea is possible. Once you've proven cyber phrenology is irrefutably real, I can unleash unholy hell at the financial cores of social justice patronage.

For those interested in the previous discussions, I'm also writing a micro-bioghraphy about how the new emotion thought experiment came to be and why that concept is essential for our efforts at reforming political theory, economics, and Western civilization.

Godspeed, anons.

Their eyes are evil and they can never smile a honest smile

We definitely need the autism.
and
So instead of using faces, we can use textual analysis? In that case, I will need to know which features should be extracted from the texts. TF-IDF keyword extraction, stylometry/writeprints, foriegn languages, emojis…
Also check this archived thread archive.fo/813BB
aicbt.com/authorship-attribution/
dariah-de.github.io/DARIAH-DKPro-Wrapper/tutorial.html
kaggle.com/christopher22/stylometry-identify-authors-by-sentence-structure
cligs.hypotheses.org/577

VR glasses, DMV data, the works. Anything that gets the job done is what we are aiming for.

So OpenFace it shall be. Still, are there alternatives to DLib?

Well someone has to gather data for it.

It is not going to help, because anything that is not owned by /ourguys/ can screw us over royally.

Is this a honeypot?

archive.fo everything please

Not sure. Either way, just use Tor+VPN to visit the Twitter.
This guy's been here a few times though. I can't remember the specifics, but he basically wrote about leveraging the reality of cyber-phrenology to morally splinter the SJW's working in Silicon Valley.
That particular idea is plausible - I do however fear that he's failed to take into account the culture of "white blame" that's been propagandized over the past few years. Even if cyber-phrenology is real, actual statistics can still be easily butchered/ignored.
Would elaborating on this spoil your gameplan?

How about using OnionShare?

As a general rule, real German names are place names or trade names, while Jew names are two random German words stuck together for no particular reason: Rothschild ("red shield") Goldbaum ("gold tree") Silverberg ("silver mountain")

regex jewish surname when?

The purpose of proving that cyber phrenology is real is multi-faceted.

1.) It allows social justice advocates to target their patronage source for virtue signalling points.

2.) It causes even the deepest Silicon Valley social justice zealot a crisis of conscience, planting the seed to turn them into a future Snowden and reveal vital secrets about Silicon Valley collusion.

3.) It forces Democratic lawmakers to intentionally cover up and turn a blind eye to the massive civil rights violations Silicon Valley is engaging in.

4.) It gives Trump the ability to pull an Eisenhower and deploy federal government assets to combat this particular civil rights violations at a time when the California Senate bid is at hand.

My efforts are not to win over the normies. It is to poison as many social justice wells as possible as close to their financial core as possible.

The problem with identifying jews by surname is that they get it from the mother (and not get the name). It is realistic that there may be a jew called "James Cristopher", and a goyim with a -berg name (like PewDiePie).

You sneaky Maplekike. Trying to get good anons to buy what is probably the worst camo ever devised by man.
This shit only works in a lab.
Get yourself some Flecktarn or Russian derivative thereof.

The best serious way to go about this would probably be to create a generic classifier that attempts to determine your heritage from a picture of your face. You need a large dataset of goyim to distinguish from kikes anyway, so you might as well go about it in a way that would allow you to operate in plain sight and make it easier to get access to the relevant datasets and harder to shut down than explicitly making a jewdar app; hell, idiots will upload their own face pics and self-reported heritage for you out of narcissism. Then, when Holla Forums inevitably starts using it as jewdar, you throw your hands up in their air and say "oh no, how horrible! we condemn this behavior!" but ultimately do nothing.

Regardless, I wonder if there are any public useful datasets out there for training this kind of thing. Doubt it but it would save a load of time; identifying millions of pictures of kikes and goyim by hand to build the training dataset is the hard part. The best shortcut I can think of would be to streamline this workflow:

You could start with celebrities and businessmen and work your way down 'til you're dealing with random low status lawyers, journalists, doctors, small businessmen, and so on. Basically anyone who has to have some kind of social media presence as part of their job, or local advertising to get clients.

kairos.com/diversity-recognition
The plan regarding the "generic classifier company"would work if Kairos does not exist.
And Kairos are made to be liberal, so I don't know if you should use their API directly.

Nice idea on data gathering though.

The hypothesis has potential, but in order for it to be better than random for generic classification, your training set will need tens of thousands of representative photos–good quality, good certainty of classification, similar shot angles, etc.
Assuming you don't end up with an overtrained pile of shit, it won't be able to make classifications of photos that aren't of similarly decent quality.
It's a cute idea, but hugely impractical if you don't have access to the data you realistically need to test the hypothesis.

This is a Jew.

Dubs of truth, the problem is that there are no dataset for us to use, we are on out own.
People are saying that VK allows country filters, so there is a way to datamine certain regions.
Iceland vk.com/search?c[country]=86 and Israel vk.com/search?c[country]=8
is basically what some people suggested. github.com/vanyasem/VK-Scraper
Regarding front shots and side shots, as said it, there are ways to mitigate such issues.

You're going to get false positives out the ass when you attempt to classify white Americans, Canadians, or Australians. Israelis also, on average, look nothing like American Jews.
Not in a way that won't meaningfully detract from the quality your training dataset.
I can tell you've never done ML in a professional setting requiring actual results.

If you can produce a dataset containing pertinent, clearly identifiable data to a human, with enough training a machine learning framework WILL find weight and work its way down the gradient through a decent artificial interpretation, computation and to a result. When you say 'instead of using faces', I'm pretty sure the main point of what is being described is the identification of faces from images, so I don't think replacing it would be very helpful. When I say to combine qualitative data, I am speaking about building up an arsenal of different tools used in tandem, each AI trained so that, when brought together, would provide multiple, accurate indices. So being trained with some thousands of images with a ton of positives and a similar ton of negatives used in tandem with a list of names with its own trained weights, and as you say, keyword extractions, stylometry, language usage or even grammatical quirks (you sure could spot a Russian/slav or German based on grammar and determiner ommission). The sky is the limit, it's all about how rich your datasets are, and how you bring them together.

AI isn't nearly as inaccessible as it is suggested by the media, and I think most of us should be putting an effort into learning it. This is the single most important tool of the next few decades and there are far too many lefties in the field.

right, the tools are out there and free to use. access to the technology hasn't ever been the major limiting factor for AI. it's 1.) getting a good set of quality data that can actually model what you're trying to model and 2.) properly interpreting the results. Just reading through this thread makes it painfully obvious that probably no one here can do #2, and #1 is not a trivial undertaking to begin with even if you're qualified to do it, let alone crowdsourced anons.

What I'm guessing is that the OP isnt actually interested in doing anything real. he's one of those narcissistic types who gets off on the idea of epeen and being at the center of some kind of movement, so if he gets a lot of attention for something he will consider that a success regardless of whether it actually does anything meaningful. there's a word for that:

Jewish.

Are you including me in that? I also know a half dozen more-than-capable regulars on here.

Actually, it's trivial for unqualified people if they are told what to do - collate lists or collections of -anything- that share a common denominator. It's so trivial that even Captcha is a great success, and that is/was with Chinese bots faking a chunk of the entries with their own tools and Africans and Indians inputting another mass of the total sample.

Save the blackpills for another topic, there's a reason it is stickied.

kvoa.com/story/37343298/monkey-see-monkey-2-scientists-clone-monkeys-using-technique-that-created-dolly-the-sheep

archive.fo/JZw8D

The team tweaked the SCNT procedure using new technology that helped with the nucleus transfer and the fusion of cells. The team spent three years perfecting the delicate procedure. Speed while performing the procedure helped, they learned, and scientists discovered clones created out of cells from fetal tissue did better than when they used adult cells.
"By optimizing the method, we obtained 79 well-developed embryos and implanted them in 21 female monkey surrogates," Poo said. That resulted in six pregnancies and the births of the two monkeys in late 2017. They hope more monkey babies will be born soon from these embryos.

————

WE MUST CREATE THE UEBERMENSCH CLONE ARMY

whoops torpedo, looks like you goofed up

We are in seriously insane times with insane news and this kind of complete bullshit is being stickied on pol.

I think there are some good mods, but there seems to be maybe one or two who are working for the other side and deliberately anchor good threads and pin dogshit stuff like this one.

This site is too important to be ruined by fuckhead (probably traitor cabal-supporter) worms who post worthless shit threads like this and the fuckhead (probably traitor cabal-supporter) mods who pin them when major, MAJOR happenings are afoot.

Fuck all traitor cabal worm fucks and all those who enable them, including the few but very noticeable bad apple mods on this site.

Get rid of dogshit threads like this. This shit has to stop.

^ CIA

[citation needed]

Jewdar googles might be useful as a learning tool. While it is striking how many jew one can identify after learning what to look for, the problem is it will dificult to spot all jews by look alone. That is eventually leading to a false sense of security, open us to infiltrators. The jews do open gloat about it, for example in the film Exodus:

Exodus (1960)
Paul Newman: Ari Ben Canaan
[Jewish Ari Ben Canaan is masquerading as a gentile British officer named Bowen] 
Maj. Caldwell : [about Jews]  They look funny too. I can spot one a mile away.
Ari Ben Canaan : [pointing at his eye]  Would you mind looking into my eye, sir? It feels like a cinder.
Maj. Caldwell : Yes, certainly.
[Caldwell looks into Ben Canaan's eye] 
Maj. Caldwell : You know, a lot of them try to hide under gentile names, but one look at their face, you just know.
Ari Ben Canaan : With a little experience, you can even smell them out.
Maj. Caldwell : I'm sorry Bowen, I can't find a thing.

imdb.com/title/tt0053804/characters/nm0000056

So they gloat about the fact that not all of them look like the hunchback of Notre Dame.

Would be interesting to build an AI which can spot the better hidden ones.

what if the Jew has a really butchered Nose Job?

Those Europeans will likely look more like a Habsburg (Prognathism, curved nose but pointing upwards, diamond face shape, small eyes and bad teeth). You can probably find a lot of those "malnutritioned, inbred" Europeans in Eastern Europe, Dixieland or Latin America.

(1) Fuck you for insulting Europeans in any way. You need to die.

(2) This thread is garbage and filler and yet another example of whatever bad mod out there pinning worthless shit when there are major things happening.

Bullshit like this should not be pinned. it is a complete, total fucking waste when MAJOR stuff is happening.

And again, a hearty fuck you to anyone who insults Europeans for any reason. If you insult Europeans for any reason whatsoever, as you just did, you deserve to die.

...

it's all about the training data, more we have more accurate it will become

Very interesting project btw, it's really fun to work with this.

I got my hands full right now as in this very moment, but I'll give this a run too once I got some free hands here. I've only worked a bit with images during school I mostly deal with boring numbers only. That means I just need to brush up on this too. It will be cool

As many others here have pointed out, if you are not into this. It's something everybody can learn btw might sound frightening but it's not.
Anyways the more data we got, the better it will get so if you can find verified (((juden))) and collect them in a zip, you are helping a lot

Do we even know that google captchas are giving them good results?

WHY IS THIS KIND OF BULLSHIT STICKIED WHEN THERE ARE ACTUAL HAPPENINGS

IS THIS SITE RUN BY DIVERTING SHILL FUCKS. WHAT THE FUCK

go make your own board kike. We don't want (you) here.
>>>Holla Forums

Who the fuck is "we" you fucking jerkoff? You have a mouse in your pocket?

Eat shit and die. I know you glow in the darks are trying to ruin this place. Trump needs to arrest every single one of you.

Regardless of how much merit the basic idea has, I agree with the other user that it's worthwhile for tech-minded people to learn about AI. Even if the whole project turned out to be a flop, people can learn from it.
I think there are some good pointers itt for that. And like another user said, companies like FB are researching this very type of thing.
It doesn't hurt to learn about it.

All AI becomes racist, and if they try to program an AI to not be racist it just won't work

When I say mitigation, I meant cleaning the dataset and removing false positives and negatives.

So basically multi-factor analysis. I like that.

Yes it is, that is why OCR and word captcha breakers are getting so good.

Depends on many factors, what text characteristics do they have, what sentiment do they have etc.

>>>Holla Forums

Why would the Jewish owner and the Muslim mod of Holla Forums put a swastika in the page?

Why do White Nationalists want to kill White people who wish to meet up without them?
8ch.net/polmeta/res/28806.html

Why would Holla Forums ban a White meet up?

Why are Whites the only ethnic group that do not meet up and network?

Jewish networking
archive.is/cdyCI

Black networking
archive.is/H6Q6e

Hispanic Networking
archive.is/SeJ8O

Asian Networking
archive.is/GTWOO

Indian Networking
archive.is/FfPFp

Muslim Networking
archive.is/rPIj1

And let's add Gay and Women networking.

vs.

White Networking
archive.is/LEhxc

Remove the Holla Forums mod
Censorship on/pol/ is worse than any of the social media sites.

STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP

There is a war against Whites being currently waged. It's time to start fighting back.

White Nationalism is the cause of "White Genocide"
123

What major things did I miss?

...

Insulting Europeans is a vice.
Well the "redstick" and "poster" thread didn't get stickied. I would rather let them have their project prioritized over this.

>>>/reddit/

Stop.

Only illegal to sell CADPAT uniforms with the IR suppressive coating intact, otherwise it's just a colored garment.

I'm a quarter Yid and my brow ridge goes straight

I wonder who they could be an allegory for…

It's like you retards have never heard of lidar, let alone facial recognition by bouncing wifi signals, or even muons.

Can quad user give an explanation to how lidar works or how they could use muons? Also I thought wi-fi signals only gave out blurry images so it wouldn't be that useful for precise facial recognition, would it?

wifi gives blurry images to those who use poor algorithms to reconstruct the image.
A 3d lidar can record the grooves on your face at a scary precision. Just add an invisible to the naked eye laser, and you won't even notice it.
Muon detectors already exist, usually used to detect uranium in shipping crates, but expect any smart house of the future to contain them. Hell, the Chinese already spy on their citizens using the mandatory housing forced on their poor.

All you really need to make a decent muon detector is aluminum pipes and some wires.

FUCK THE MOD WHO STICKIES WORTHLESS WASTE OF TIME SHIT LIKE THIS THREAD THAT DOES NOT REMOTELY HARM THE KIKES. THERE IS SOMETHING SERIOUSLY WRONG WTH AT LEAST ONE MOD HERE WHO STICKIES BULLSHIT LIKE THIS BUT NOT GODDAMN WORLD WAR III ISRAEL IS STARTING. THEY ARE TRYING TO START WORLD WAR III.

FUCK YOU MODS FUCK YOU MODS

Calm.your bitch tits

I am talking about the general ethos of Holla Forums. Sure I might be out of line but kvetching like a kike is a clear sign that (you) don't understand how infinite chan operates. Fuck off, go make your own board.

Some resources from the last Antifa facial database
notabug.org/JerryShaw/EyeOfAnEagle

To think a sticky died for this. Didn't Uncle Adolf have an entire fucking picture book on how to spot kikes?

Is it in >>>/zundel/ or >>>/pdfs/?

Re: Ivanka Trump's children are Jewish.

just provide dox and let the right wing death squads finish it

I wish I knew any coding to be able to help but I'm not that kind of stem. If you need numbers to math I'm your guy though. Really wanna see this made. Gonna use it on my own face, for reals

What benefit would that have?

enjoy

instructables.com/id/The-Invisible-Face-Mask-/

No computer will ever match the intelligence of a fly you techfags.

Intelligence is, literally, the absolute product of biconditional transmission within an interneural framework undergoing n level of activity, where n =/= 0 - itself conditionally existent upon structures of grey matter subject to a sufficient level of plasticity. Therefore, under the law of bivalence, ergo employing proof by contradiction, given that a fly does not contain within its structure the presence of grey matter, ergo the capacity for biconditional transmission, and a computer similarly has no such capacity, a computer shares, literally, the same level of intelligence as a fly. The computer, however, still contains within it the potential for biconditional transmission, at a minimum under simulatory conditions, which have already been produced by almost every university laboratory in the West and Asia, and to a degree where the presence of an active intelligence can be debated.

So stop talking like you want to get DOTR'd.

This is exactly the core of the task. We need a high-quality dataset - by high-quality of course I mean properly labelled dataset, and the photos have to be of the same type and quality as those you plan on processing with the trained system. So super-high pixel count and well-lit pictures for training set is complete garbage if what you want to classify in the wild is poorly lit cellphone camera images. The training set must come from the same sources that you would wish to later classify.

Also, Jew/Not-Jew may turn out to be … unrealistic? I'm not convinced that there are any real identifying physical characteristics that are common across jews. The jewish race I think is essentially a lie. Look back through the primary source historical documents.

I actually think you may have far better success processing voice data.

Kill yourself kike.

It would be useful to have a system to experiment with that uses either Caffe or Tensorflow.

developer.movidius.com/

These development USB sticks are about $60 and will return a result in ~150ms from the trained system. The chip itself costs much less and can be integrated as an AI accelerator into whatever.

Who has the PDFs from a similar thread on this subject from a year or two ago?

Remember this old thread, Recognizing Jews?"
Where can I get these old pictures, like the one shown here?

I never really understood this method of detection. People aren't made of Lego blocks, everyone's brows curve downward into their noses. Pic related, the lines could have just as easily been drawn the other way around.

It's said this for a long time and finally someone on else noticed it.
I will also draw a picture and label it very simply even though pozzed/cuck/commie is overused but maybe you get my meaning. Just take a look at 100 pictures of normal people and those that put their neck out too much or are liberal crybabies. You'll notice it too eventually.

Look at the brow, it follows the nose down.

I don't know about cucks, but there's always that one fucking face coupled with a slightly drooping jaw. Of someone who you can tell is just a useless idiot. No goals or anything in life. Even the reddit fedora tipper pictures never are as retarded looking.

Lenin was a loxist who got away with killing millions of whites, he was a commie and the enemy of civilization but he wasn't a useful idiot or a pozzed cuck. This goes for all loxist jews, but it's the pieces of shit that play on their side not knowing it's the antithesis of what we would be striving for that have the outward brow most of the time.

Provide examples, desu.

wax

Speak for yourself. I know how to do that part just fine, but there's not a lot to do unless someone can put together a proper dataset.

An ideal dataset would also come with a DNA map, if we're talking wishlists. A set of faces in various conditions, associated with their DNA profiles. I suspect that you could probably get good results out of something like that, especially if you turned it into a phone app, bill it as 23&Me 2: Smartphone Boogaloo. I think you'd do better than giving a single readout saying "76.3% chance of JEW", especially given the mish-mash of genes these Jewish mutts are as someone else noted. If someone produced something like this, I think they'd probably make a lot of money.

a computer could match the intelligence of a fly, the difficulty would be in giving it a fly body that was worth a shit.

What the fuck.

please check out the enemy has already made a version of this, and they are the ones with massive resources

This shit is ridiculous fantasy distraction when our entire government is treasonous, they are passing Stasi-style left-wing death squad legislation, and Jews are trying to start WWIII and cause nuclear armageddon.

Reminder that the Israel starting WWIII thread wasn't pinned until I protested.

Shit like this thread is total shit distraction, and whoever is responsible for "pinning" this when the world is burning down is an insider who's trying to destroy pol. They need to go.

But user, that's crazy talk. To think that Holla Forums is moderated by nonwhite kikes who are personally protected by jim because he has a false narrative to spread is obviously not true. Ignore all the proof of this.

Don't be an idiot. If they had any say so on this site it would be like 4chan. Shill posts wouldn't be removed when we report them, and any threads with real info would just vanish.

Uh… that's exactly what fucking happens here, you idiot. Why the fuck do you think PPH has dropped by a factor of twenty in just the last two years alone? Because hotwheels is dead, jim's a kike, and his self-appointed, personally protected nonwhite moderator has deleted anything that goes against Trump's zionist civic nationalist plan.

you are actually retarded. What the fuck do you think "thinking" is? Some supernatural spiritual thing? No, its a material process done by our brain computer. AI will eventually be able to emulate any sort of pattern, even human patterns. Your hippy humancentric nonsense is as retarded as saying that the sun revolves around the earth.

Provide evidence for that claim.

Excellent post user.

Why should I believe anything? Also the jews want mulattoes so white women are brainwashed in to fucking leroy not being poor and sold to foreign powers.

>>>/judaism/

White Nationalism/Alt-Right is a Controlled Opposition Jewish psych-op whose sole purpose is to leave Whites defenseless as they slowly murder us.

OK. where go then?

If you want this to work you need a representative sample. You need to gather pictures the way this thing is likely to be fed them and then label them. It is critically important that the sample has the same proportion of Jews to non-jew as it is going to encounter in the wild. This is what is know as a prior probability and is one of the most important metrics used by these systems.

It will not work if you just give it a 50/50 mix.

Just pointing out "in the wild" is context dependent. Looking for jews in movies + news is going to give you a much different ratio than scraping facebook or streaming a street camera. So a use case will need to be decided on before training the system.

This is useful and it should be install in the drone along with the dna scanners.

My concern is that Jews will try to inevitably try to spin this as phrenology or a pseudo-science

How do you counter this as a talking point among friends and normalfags?

(You) Beat me to it.

According to that pic the first guy in the third pic in is a cuck.
i.e., it's nonsense.

DNA testing and biometrics information database need to be mandatory for everyone so there is no privacy for the jews to hide from.

...

They should be more concerned of the ones government agencies will develop in an attempt to extract as much information about your general health, age, social status, ethnicity, emotional state and so on with generalized facial and body recognition technology in development. Advanced enough AI in such recognition will be able to pick up subtleties of a face or body that humans take for granted or are only subconsciously aware of. You could say AIs would practice an "instinctual" behavior like how people are able to accurately read someone just by casual glance.

It'll become increasingly difficult to argue that point of such technology being pseudoscience. After all, the appearance and condition of your flesh would reflect on a lot of your behavior from personality to diet. So why would easily identifying ethnicity with such precision be any different or controversial at all? That the question I'd ask them.

Besides I wouldn't put it past the Chinese that they already have (or will have in a few years) something this advanced and have linked it to their social credit system.

Learn OPsec faggot, and open it in a VM.

Is OP dead yet?

White Nationalism/Alt-Right is a Controlled Opposition Jewish psych-op whose sole purpose is to leave Whites defenseless as they slowly murder us.
111

He's probably some faggot that went OH NO STOP THIS IS TOO RACIC4ME.

You guys understand that the nazis actually failed miserably when they attempted to identify ethnicity by phenotype, right?.. Guys?.. Do you even understand the word phenotype?

What you guys think of the alien hoax the Q-user is slowly pushing? Is the military communist alien hoax a threat to christianity because of the communist factor? Remember USSR history and christianity. They claim Anunaki is the real story and the Holy Bible is a corrupted fabrication. This meant they need to get rid of christians, jews, muslims which make up billions of people (Agenda21).

I found this Q-user message on a sub board:

Coincidence the Matrix (movie) grew people as a crop, used for energy, and controlled their mind?
Sound familiar?
Wonder where they derived that idea from.
Now comes the 'conspiracy' label.
Deeper we go, the more unrealistic it all becomes.
The end won't be for everyone.
That choice, to know, will be yours.

Q

NSA confirmed. Q is going for "the end won't be for everyone". He refers to the Annunaki alien psyops which is enslaving humankind (and they need the Army and NSA to do it on their behalf…)

This was their plan all along since 1947

There is no cabal replacement. They will end the FED or fiat currency and introduce the cashless society giving total control over You to them…. the army… I mean the aliens.

They will actually kill the disbelievers and create a revisionist / alternate history.

Q is fake

Won't stop people or organizations from trying, especially with access with current and emerging AI technologies.

We first recognize people by their race by how they look. If you care about phenotypes instead of just what you see with your eyes you're advocating for white genocide.
I should have known.

TFW you finally realize why so many kikes wear glasses.

Apparently Hitler was also quarter yid.

Although given how politicized he is at this point I consider whether that's true a coin toss. But that's the general accepted fact of the matter.

They.

No it is not, kike. OUT

Time to calibrate.

They are more people than you are, Tyrone.

You can't just identify someone's ethnicity goyim.

Why would the Jewish owner and the Muslim mod of Holla Forums put a swastika in the page?

Why do White Nationalists want to kill White people who wish to meet up without them?
8ch.net/polmeta/res/28806.html

Why would Holla Forums ban a White meet up?

Why are Whites the only ethnic group that do not meet up and network?

Google:

Jewish networking
Black networking
Hispanic Networking
Asian Networking
Indian Networking
Muslim Networking

And let's add Gay and Women networking.

vs.

White Networking

Remove the Holla Forums mod
Censorship on/pol/ is worse than any of the social media sites.

STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP
STOP-POLL-CENSORSHIP

There is a war against Whites being currently waged. It's time to start fighting back.

White Nationalism is the cause of "White Genocide"
12wqa

Forensic anthropologists can identify the race of a corpse from bits and pieces of the skull.
You leftist cucks who think your delusional beliefs are based on science are as dumb as mudslims who claim science proves the koran is true

D E L U S I O N A L

ANOTHER (1)
Anyone who brings up Hitler was part jewish is controlled by jewish propaganda. Do these fags really ever think for themselves? We need to keep in mind that UK and US psyop groups are still lurking and posting here so it's quite possible they try to test this place with (((out of africa))) and (((hitler was part jew))) mainstream lies.

He had one job.

this is the stupidest fucking shit i have ever heard.

Hitler couldn't have done shit on his own without the help of the Zionist party. after the war was ending the zionist threw hitler under the bus and basically won the war. the zionist wanted to punish orthodox jews who believed that jews couldn't create a nation until their messiah arrived. the zionist jews believed that they can create their own country. the zionist punnished the orthodox jews by having the nazi's round them up and put them in concentration camps, the concentration camps were ran by Zionist jews. all the orthodox jews that wouldn't submit were killed and burned, the ones that would submit were to be shipped to the middle east after the war to buy their own country. Israel would literally erect a statue of Hitler as the father of Israel if it didnt arouse suspicion and basically destroy all the lies they have told in history.

No. That will not work the way you think it will.

The nose knows.

AR glasses are shit, just have it apply to video and images. You would be forced to have a really shitty low-resolution camera, low refresh rate (which is made even worse by the fact that you are analysing in real time), fuck-all battery and a display that is still garbage even after putting it right in front of your eyes. There is a very good reason why AR glasses aren't ever used. Just deal with it analysing raw footage.

Except it isn't. The original iteration of this thread happened weeks to months ago, and it wisely included a link to a study.
This study found that by using face scanning A.I., researchers were able to determine a common theme between facial structure and sexual orientation on a dating website.
Not only that, said A.I. was then able to predict a person's sexual preference using that data with the insane accuracy of something like 90%.

Well if we can predict something as nebulous as sexual preference by looking at someone's face, then surely it would be even easier to determine a person Jewishness. Especially given it's obvious genetic basis.
With a large enough reference database, using only a single photo you could determine with a high level of accuracy if someone is a crypto-kike.

tl;dr you must be a literal kike if you can't appreciate the implications of this

Bump

Hope it helps:
github.com/AyaUberrheiner

includes:
- facial recognition and alignment tool
- Image scraper from Google image searches (up to 1K/query).
- Pedo classifier.

Google did some weird shit to their image searches in the last few days. I haven't retested. You may need to modify the script to grab thumbnails.

The basic form of the pedo classifier should be applicable to this problem.

>github.com/AyaUberrheiner
Let me get this right, you got something that can tell a pedo from a non-pedo with 75% accuracy? Neat.

archive this

Love the project, hate that they have a folder for "White" males.

Do it yourself.

my.mixtape.moe/kagtex.pdf
my.mixtape.moe/ijndrp.pdf

k sperg

Is this loss? (pun intended)

Seriously though, are we still spinning around talking about whether there are datasets or not?
Because we should sort this out somehow.

It would be useful like that

Found what you are talking about. There is a tutorial literally
medium.com/@ageitgey/machine-learning-is-fun-part-4-modern-face-recognition-with-deep-learning-c3cffc121d78

good to have a continuum instead of a pure yes or no. if you can have the corpus be (Name, Picture, PercentJewish) that'd be great

"Select all images matching the description: jew"

who wants to use twitter APIs to grab a list of twitter users who flies the israeli flag? Or better yet, geo-tagged users?

the old way was best but this is next best thing

Trump says take the guns. No Due Process
Trump says take the guns. No Due Process
Trump says take the guns. No Due Process

stumbled on a pure specimen

I agree that it is our magnum opus. Think about it. Think of how much kvetching will they do when the Jewdar is complete and is demonstrated to work. Guaranteed to be talked about by MSM immediately if we do it right. I’ll be monitoring the threads to see if I can help.

But none of us have made a dataset, that is the probelm. We need to scrape some jewish schools and Israeli social media to get those faces.

bump

Cryptojews and false positives aren't a huge deal, I've been working on this off and on. I'm no expert but my current thinking is an adversarial autoencoder. We don't need labeled data, we train it up on whatever we can and then feed it known kikes and see how they're encoded - I expect they'll cluster. Bonus: we can generate faces to visualize what the network has learned.

holy shit

Do some research on Binary GANs. You need data and classes to train.

Attached: 0_MPxewjAGEQpfls3W_.png (950x420, 100.98K)