But this politician said «xyz»! - no Commentary
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Subtitle |
| |
| Title of Series | ||
| Number of Parts | 275 | |
| Author | ||
| License | CC Attribution 4.0 International: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/52350 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
| |
| Keywords |
00:00
Electronic visual displayHypothesisTwitterChaos (cosmogony)VideoconferencingArtificial neural networkHash functionOvalLevel (video gaming)MassGreen's function
02:48
Home pageSoftware testingSoftware engineeringPhase transitionComputer-assisted translationMultilaterationCoefficient of determinationSoftwareMultiplication signSystem callCASE <Informatik>19 (number)Video gameWeb pagePhysical systemComa BerenicesComputer animation
03:48
Computer networkCASE <Informatik>Negative numberSoftwareComputer animation
04:11
Exploit (computer security)VideoconferencingPlanningEvoluteHypercubeStudent's t-testFigurate numberSoftwareVirtual machineMultiplication signComputer programForm (programming)Artificial neural networkNeuroinformatikMachine learningDisk read-and-write head
05:08
CAN busWeightComputer-generated imageryVideoconferencingArtificial neural networkCartesian coordinate systemNoise (electronics)MereologyPhysical systemVideoconferencingImage resolutionElectric generatorFunction (mathematics)SoftwareInformationVirtual machineFeedbackMultiplication signForm (programming)Computer-assisted translationTouch typingArtificial neural networkoutputLine (geometry)PixelPattern languageProcess (computing)WordCASE <Informatik>Reduction of orderQuicksortShape (magazine)Point (geometry)Real numberObject (grammar)Sampling (statistics)Entropie <Informationstheorie>Order (biology)HoaxState of matterLevel (video gaming)Video gameMatching (graph theory)Office suiteCellular automatonCasting (performing arts)Shared memoryWeightLattice (order)TouchscreenConnected spaceIntelligent NetworkComputer animation
12:00
Game theoryLevel (video gaming)VideoconferencingForm (programming)Video gameLevel (video gaming)Metropolitan area networkAlgorithmDifferent (Kate Ryan album)Game theorySoftwareTerm (mathematics)WebsiteCellular automatonHoaxOpen setComputer animation
13:14
Cartesian coordinate systemComputer-generated imageryYouTubeTerm (mathematics)SoftwareRight angleSource codeMeeting/Interview
13:53
Multiplication signNear-ringRoboticsNeuroinformatikDecimalYouTubeSource code
14:18
Cartesian coordinate systemComputer animationLecture/ConferenceMeeting/Interview
14:39
FreewareDigital filterSocial softwareGame theoryFilter <Stochastik>HypermediaVideo gameFacebookComputing platformKeyboard shortcutCASE <Informatik>Computer-assisted translationMultiplication signSoftwareSurjective functionVideoconferencingCoefficient of determinationMeeting/InterviewComputer animation
15:39
CASE <Informatik>Multiplication signSystem callPosition operatorPixelRevision controlAlgorithm
16:12
AlgorithmRevision controlSimilarity (geometry)SoftwareTwitterVariety (linguistics)View (database)Term (mathematics)Graph coloringMeeting/Interview
17:13
FacebookDecision theoryIdentity managementHoaxAlgorithmFacebookHypermediaPoint (geometry)Core dumpRobotTwitterCodeComputer animation
18:04
Cartesian coordinate systemStreaming mediaComputer-generated imageryVideoconferencingElectronic program guidePlot (narrative)HoaxHypermediaProcess (computing)Mobile app
19:18
Social softwareBeta functionIdentity managementPort scannerHoaxSoftwareSoftware development kit
19:50
Power (physics)HoaxMeeting/Interview
20:10
Order (biology)Asynchronous Transfer ModeComputer-generated imageryCartesian coordinate systemHoaxLine (geometry)Arithmetic meanMeeting/Interview
20:32
Cartesian coordinate systemSet (mathematics)AlgorithmNumberWeightInsertion lossTouch typingVirtual machineInformationChemical equationPoint (geometry)Open sourceError messageDot productGraph coloringHoaxCASE <Informatik>Endliche ModelltheorieGreatest elementArmPhysical systemTask (computing)DataflowVariety (linguistics)Bounded variationNegative numberGroup action
22:27
Cartesian coordinate systemOpen sourceExplosionComputer-generated imageryCAN busOpen sourceAlgorithmTraffic reportingHoaxComputer programPower (physics)Product (business)Computer animation
22:57
Power (physics)Computer programDependent and independent variablesMeeting/Interview
23:20
Multiplication signRevision controlDigital photography2 (number)Row (database)outputLatent heatNeuroinformatikCASE <Informatik>Meeting/Interview
Transcript: English(auto-generated)
00:17
and welcome back to the R3S of the RC3 in Monheim.
00:28
One thing that came up in the IRC that was unrelated to any talk was about the little display you see here next to the Winckekätze.
00:43
What this is, it is a CO2 indicator, so we can see when we have to ventilate the room. So, one of our producer, the nice guy who does the video here, bought this
01:12
so we can see when we have to exchange air to prevent the spread of aerosols and so on.
01:23
Anyway, if you have any questions regarding our stage or the talk, please feel free to join the IRC channel rc3-r3s on Hackind IRC
01:46
or use the hashtag rc3r3s on Twitter and Mastodon or use our handle at r3s at chaos.social on Mastodon.
02:02
So, in our next talk, we're going to stay with artificial intelligence and with GANs. Now in English, as you may have noticed. Our next speaker normally haunts the hackerspace in Ghent.
02:27
She does a master's thesis on GANs. She's very interested in the ethical aspect of what this technology can do. So, please have a very warm welcome for Lisa Greenspax and her talk,
02:46
but this politician said XYZ. Hello and welcome to my talk, but this politician said XYZ. I want to talk today about the technology behind deepfakes and its ethical implications. So, if this was a live talk, I would have asked you to raise your hand if you knew this person
03:03
and I would have expected nobody to raise their hand and if they did, I wouldn't believe them because this person does not actually exist. Thispersondoesnotexist.com is a home page launched in February 2019 by a software engineer at Uber and every time that you refresh the page, another phase shows up
03:21
which was generated by the same technology that is behind the deepfakes, which are called the generative adversarial networks. Next to thispersondoesnotexist, there is also thiscatdoesnotexist.com. So, you could argue that this cat does and does not exist at the same time. To my disappointment, this dog does not exist yet.
03:44
So, maybe that's a test for later. So, let me walk you briefly through what I'm going to talk about today. So, first, I'm going to explain to you what generative adversarial networks actually are. Second, I want to give some use cases, what GANs are already used for
04:01
and then we're talking about the downsides of GANs. So, for example, deepfakes, but also other negative use cases of deepfakes. So, what are GANs? GANs were introduced by Ian Goodfellow in 2014. Ian Goodfellow is an ex-Googler and now is the head of machine learning at Apple.
04:22
He's also a former PhD student of Andrew Ng, who's a very popular figure in deep learning. And in this street here on the right, you can see a post of Ian Goodfellow in 2019 about the evolution of GANs. So, we started on the left with a very pixelated black and white picture
04:42
of a woman in 2014. We go through the years up to 2018, where we already have a very photorealistic picture of a person that does not exist, generated by a computer program. Now, as you have seen, this person does not exist now. We even have hyper-realistic pictures of people
05:02
that we can't even distinguish from real people anymore. So, what are GANs? GANs are short for generative adversarial networks and GANs consist of two neural networks competing against each other.
05:21
So, on the one side, we have the generator that generates an image or audio or video, for example, and is also sometimes called the artist. The discriminator, on the other hand, discriminates an image or video and audio. It's also called the art critic. So, it's telling whether an image or whatever other input is realistic or not.
05:43
So, that's a lot of new words. So, let me walk you through them. So, what is a network, actually? As a disclaimer, this is very simplified for you. So, please, fellow machine learning engineers, don't touch me on that. So, a neural network is based on the idea of human-brain physiology.
06:04
And each node in that neural network would be a neuron in the human brain, connected to other neurons, forwarding and transforming information. Neural networks are a part of deep learning, which is, for example, part of the hidden layers in the neural network.
06:22
And deep learning is a part of machine learning, which is a part of artificial intelligence, mimicking the human brain's intelligence. So, a neural network typically consists of three main parts. We have the input layer, we have one or more hidden layers, and we have the output layers.
06:41
So, for example, our input layer could be an image of a cat. That would be the RGB values, the pixel values of the cat, which are getting forwarded to the one or more hidden layers. And the hidden layers are doing some sort of feature extraction.
07:02
So, simplified, you could say that the hidden layers are checking whether there are pointed ears or a cute nose with screws or the typical eye shape of a cat, and it's getting this kind of information forward to the output layer, which calculates a probability,
07:21
how likely it is that the image that we put in is actually the image of a cat, yes or no. And this is basically what our discriminator is doing. Our discriminator is the art critic that sees an image, for example, of a cat, and it's supposed to tell us
07:40
whether it is indeed the image of a cat or not. Our generator, on the other hand, works the other way around. So, it gets a so-called noise as an input layer. Noise are randomly sampled values. It forwards those to hidden layers,
08:02
which are supposed to form ears, eyes, snout, and so on, and it transforms that into pixel values to generate the image of a cat. So, how is this working together now? The generator, who only gets random noise at the beginning, starts to draw very random stuff
08:21
that can be blobs, black and white, lines all over the place, and it forwards those generated images to the discriminator. The discriminator, who does not know yet what a cat is, then makes a guess. Is this quickly lying here a cat or not? So, in this case, let's say it gives the information,
08:44
no, this is not a cat, back to the generator. The generator then knows, oh, okay, well, I have to change something about that. So, it keeps trying and trying and trying until it gets closer to what an actual cat is supposed to look like. The discriminator is not only learning through the generator and its output,
09:04
but it's also learning by getting real images of cats. So, the discriminator is getting the fake images of the generator, but the discriminator is also getting real images from our labelled input data. So, every time the discriminator sees a picture, it makes a guess.
09:23
So, yes or no, is this a cat? And then it gets feedback from the system by, okay, this is a real image or this is a fake image generated by the generator. And the discriminator's goal is to be able to differentiate the two,
09:42
to say, okay, this is fake and this is real. And the generator's goal is to make pictures as realistically as possible. So, these are our two neural networks, the generator and the discriminator, fighting or competing against each other. So, that's the adversarial part of generative adversarial networks.
10:03
So, this process keeps going on and on until the generator can generate pictures that are indistinguishable from our real images. Some of you might have seen this rather popular GIF already
10:22
from an paper from Su et al. in 2017, where our inputs were moving horses and the input was images of zebras and the GANs goal was to map the pattern of a zebra onto a horse.
10:43
And while this looks very funky, if you took a screenshot of it, it would, to most of us, at least look like a zebra. So, now that you know what GANs actually are, what are GANs actually used for? What are they useful for? So, they are, for example, used in medicine,
11:00
for example, to reduce noise in images, so fragments that are not supposed to be there. They are there to up-sampling images. So, in case we have a low resolution that upscales to resolution, we have classification, is it this or that, we have segmentation and we have object detection. So, here in the lower left image, you see pictures of an eyeball
11:25
and a GAN, for example, could extract an image of the blood vessels in this very eyeball, which then could be used to diagnose something or to see whether everything is working fine. And on the right image, you see an MRI scan of the brain
11:44
and the GAN would be able to detect abnormalities in the tissue that could give hint to a disease or something which might not be visible to the eye or could at least save time in the process and resources. So, science is one big point where GANs are already used for,
12:03
but GANs are, for example, also used for arts or video games. So, here we can see that it has been used in The Legend of Zelda from 86 in a paper from Torato et al. from past year that a GAN could generate new levels in this game. And about 60% of the levels that the GAN generated
12:24
were actually playable levels. So, in these kind of levels, you always have to have a certain amount of items, you have to have a key, you have to have a door and so on. And GANs were able to produce up to 60% of playable levels compared to other algorithms
12:41
from which only about 10% were playable. Another form of art beside video games are movies where deepfakes or GANs are already used. So, here the Reddit user Derpfake uploaded a GIF of the face of Nicolas Cage
13:01
put onto another actress's body from the same in the movie. They were both starring in Man of Steel. Nicolas Cage's face on different bodies has gained quite some popularity in the recent years. And beside putting Nicolas Cage's face on other actress and actress's bodies,
13:20
other users have shown that generative adversarial networks can also outperform CGI, which might be used in the creations of movies in the long term. So, here you see that in the movie Rogue Run, the young Carrie Fisher on the left side with CGI and on the right side with deepfakes or GANs
13:41
produces a far more realistic and prettier picture. Another example is Robert De Niro's The Irishman,
14:01
where he was de-aged since the actor was already 70 years at the time and it took Netflix about $10 million and two years to de-age Robert De Niro while it took one YouTube user about one week and his home computer to generate this.
14:48
So, next to science, arts and video games, there is another use case that most of you people have either used themselves or at least have seen on social media platforms. And they're so-called filters.
15:01
We have aging, we have face swapping, we have putting bunny ears, cat ears, dog ears onto people's faces. They're all also created by generative adversarial networks. Another use case, for example, is Crea Therapy that has been used for the first time in this year,
15:21
where therapists have spoken with the voice and face through somebody who just passed away unexpectedly. So, for example, a father could talk to his just-passed-away daughter and work on his grief through that. So, what are the downsides of GANs?
15:42
Now, we've seen many positive use cases, many useful use cases, but GANs are not without any problems. So, one big problem is bias and especially racial bias. And here on the left side, you see a pixelated picture of former US President Barack Obama
16:01
that got up-sampled by NVIDIA's StyleGAN algorithm into a very whitened version of Barack Obama. Twitter user usasubo has used this algorithm a couple more times, where you can see here on the left side
16:20
the original images of the people that he used. So, he first pixelated them, which is the middle picture, and put them into the algorithm and the right column is what the algorithm puts out. So, here you see a variety of ethnical backgrounds and skin colours, and their pixelated versions got up-sampled
16:41
in a very whitened version of them. So, another problem is that GANs only produce pictures with bias, but similar techniques are used to predict the probability of which people who are accused of crime would commit crime again.
17:01
And this also shows a substantial racial bias with people of colour who are getting longer sentences because the judge would use such biased software. Another problem with GANs is that they can be used to create fake identities.
17:20
So, for example, social media bots are getting more and more realistic and are used to influence people's political opinions and decisions. So, Facebook removed over 900 accounts which spread pro-Trump propaganda to about 55 million users. Facebook held a coding challenge to develop an algorithm to detect fake images,
17:40
which they called the Deep Fake Detection Challenge in December 2019. And Twitter, for example, said that they're marking tweets that contain fake images and wants the user when they want to share the tweet with a fake image. Most of the algorithms that are supposed to detect fake identities or deep fakes are typically also based on GANs.
18:06
One of the biggest issues and downsides of GANs are identity theft. And about 96% of all deep fakes are porn. That is celebrity pornographic videos, but, for example, also revenge porn.
18:22
And this created a whole sub-genre of porn. So, it's mainly used for against actors and actresses, but also you and me could be a victim of this process and, for example, political opponents. So, while I was researching articles about deep fake porn and so-called deep nudes,
18:45
I found this terrible article reviewing the best deep nude apps of 2020, which I tried to report, so let's hope that it's getting removed at least. And in 2018, people tried to silence Ahrana Ayub,
19:03
who is a Muslim and investigative journalist from India. Her social media account got infiltrated with fake posts and fake porn, such that she wasn't accepted at any Indian publisher anymore and she couldn't leave her house for quite a while.
19:21
A last problem that I wanted to mention is the tampering with medical imagery. So, it starts to spread to other domains as well. So, researchers have shown that you can inject or remove a tumour on an image of a 3DCT scan of a lung that fools medical professionals, as well as detection software.
19:42
And then there are many things that we're probably even not thinking of yet that the gun and deep fakes could influence and take over. And I would say, why is that a problem? Only people who have a lot of commuting power or a lot of knowledge about these things can create deep fakes.
20:01
But that's not true. Basically, everybody can create deep fakes. Ojola ifeir. It is important for you to know that everybody can make deep fakes now. You can turn your head around. Mouth movements are looking great
20:20
and eye movements are also translated into the target footage. And of course, as we always say, two more papers down the line and it will be even better and cheaper than this. So, now I've mentioned all the dark and negative sides of GANs. But what shall we actually do?
20:41
What can you and me do against the known sides of GANs? So, when we talk about bias, especially as a researcher, there are several things that you can do. So, you need to try to balance your data sets. And you can do that with a variety of things. You can try to put more variation of collection methods.
21:02
You have to have a high diversity of people labelling your data. And you also need diversity of where your data is collected. So, here the image at the bottom, you see on the left side ImageNet, which is a very popular data set with live-built images, but more than half the data was collected in the USA and in Great Britain.
21:22
So, it's not very representative of the world, but it is used in all kinds of tasks. And it's also usually the case that the dominant culture is often higher represented in a data set, or even reflect correctness when it's put into an algorithm. On the other side, you can try to balance your algorithm.
21:41
That can be done by checking losses and weights, etc. But also the people who are coding influence the bias of an algorithm. So, there was an example with a soap dispenser where a group of people with white skinned colour developed a soap dispenser that would react to your hand
22:00
pulling under the soap dispenser, and it would not react afterwards to people with a darker skin colour. So, your bias is also reflected by the people who are coding. So, to sum up, bias can be introduced to a machine learning model, basically at any point where a person might have designed, engineered or touched the system,
22:21
and everybody of us is biased, whether they are aware of it or not. But what can we do against deepfakes? Unfortunately, not a lot, because the technology is already out there and it's already freely available for a lot of people. What you can do and what I can do is we can question our sources of information,
22:42
our sources of images, etc., to detect deepfakes. And if you're sure that you detected a fake, report it. And also don't use or support GAN algorithms in a harmful way. Thank you for your attention. I hope you found it interesting and learned something new.
23:03
I'm looking forward to answering your questions. And don't forget, whether you're programming with GANs or whether you're using or consuming GANs or their products, don't forget, with great power comes great responsibility. Thank you.
23:22
So, yeah, I think I'm back on stage. Oh, nice. So, I see that there is... Oh, OK. So, your talk seems to have been quite comprehensive
23:42
and as nobody did leave any questions. Considering that this is the second talk on GANs in a row, I think, yeah, what was really nice of you
24:01
was that you covered the other side of GANs. The speakers for you tried to cover or did cover a specific use case and how to implement them. And you covered more fundamental concepts.
24:26
So, yeah, thank you very much for taking the time of preparing and giving this talk. And, yeah, have fun on the remaining RC3. Thanks for virtually coming over.
24:43
And the... Oh, yeah, one came in. One came in just now. How big is the computational cost for a discriminator? Oh, again? How big is the computational cost for a discriminator?
25:04
For a discriminator, that really depends on how complex your GANs is. If the input for your GAN is very small images then the computational cost is, of course, well as well. You can't, yeah, you can't generalize that really. OK, so, thank you for taking your time.
25:23
Thank you for answering this last-minute question. And, yeah, have fun. Thank you for hosting me. It was really nice. You're absolutely welcome. Thank you for coming over.