Scribd
Upload
108 views2 pages

How DeepFakes Work

Deepfakes are fake videos or audio that are manipulated to make it appear that a person said or did something they did not. Deepfakes use generative adversarial networks (GANs) which involve two machine learning models - one that creates fake content and the other tries to detect it. The creator model gets better at making fakes as it trains against the detector model. Deepfakes are concerning as they could damage reputations if used to manipulate statements from political leaders or create nonconsensual pornographic content. Blockchain technology may help authenticate original media by providing a tamper-proof record of origins and manipulations.

Uploaded by

jack sparrow
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
108 views2 pages

How DeepFakes Work

Deepfakes are fake videos or audio that are manipulated to make it appear that a person said or did something they did not. Deepfakes use generative adversarial networks (GANs) which involve two machine learning models - one that creates fake content and the other tries to detect it. The creator model gets better at making fakes as it trains against the detector model. Deepfakes are concerning as they could damage reputations if used to manipulate statements from political leaders or create nonconsensual pornographic content. Blockchain technology may help authenticate original media by providing a tamper-proof record of origins and manipulations.

Uploaded by

jack sparrow
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

How DeepFakes work?

You might have come across this term trending nowadays, scrolling
through memes in your Instagram feed or Youtube. Take examples
of Obama’s public service announcement, Trump’s joining Breaking
Bad or Zuckerberg openly claiming he exploits the personal data of
Facebook users.

Deepfakes are basically fake videos or audio recordings that look


and sound just like the real thing.

This particular deepfake algorithm allows a user to take any video


of a person’s face and use it to animate a photo of someone else’s
face with only a few lines of code. This is just a few steps ahead of
the face-swap apps that surfaced few years ago.

Deepfakes use generative adversarial networks (GANs), which is


one of the most impressive outcome from AI Industry in recent
times having the potential to recognize a pedestrian from across
the road. But the most important vision it aims for is to give the
computer a hint of imagination after feasting on raw data and
hence learning from it.

GAN involves two ML Models, one model trains on a data set and
then creates video forgeries, while the other attempts to detect
the forgeries. The forger creates fakes until the other ML model
can't detect the forgery. The larger the set of training data, the
easier it is for the forger to create a believable deepfake video.
This is why videos of former presidents and celebrities have been
frequently used in this early, first generation of deepfakes —
simply because there's a ton of publicly available video footage to
train the model and hence increase our accuracy and of the
deepfake.

It may sound interesting and fun until it is used to destroy


someone’s reputation especially in case of Presidential candidates’
statements or pornographic content. It doesn’t take much time to
mentally harass someone with this kind of technology available to
public or have even adverse effects on a national level.

SO, where is the checkpoint? How do you recognize a deepfake with


the original video? One of the strategies coming up is the use of
provenance of the media. A blockchain online ledger system could hold
a tamper-proof record of videos, pictures and audio so their origins
and any manipulations can always be checked.

But we still don’t know what potential and its limits AI Industry
holds for us, in the coming years.

-Yash Verma

References:

[Link]/article/3293002/deepfake-videos-how-and-why-
[Link]

[Link]
deepfakes-and-how-can-you-spot-them

You might also like