Can You Spot the Fake Generated Face?
Here is an exercise for you:
Can you spot the fake face (or faces) among the six images below? By fake I mean generated (in other words, the person does not exist and has never existed).
See the faces in larger format below (click on the image for an even higher resolution):
If your answer was one (or more) out of A, B, C, D, E or F then you are right. Actually ALL the faces are fake and the people in the images do not exist, have never existed and most likely never will.
The faces were "imagined" and created (in high definition) by an Artificial Intelligence algorithm. Specifically it was created by a type of Deep Learning algorithm/model known as a GAN which is short for Generative Adversarial Network.
While one of the networks (the Generator) generates the fake data, the other network (the Discriminator) tries to detect whether the generated data is fake or not.
If it succeeds in detecting correctly, the Generator tries again using the feedback from the Discriminator to improve the "quality" of the generated data.
This back and forth process goes on for several cycles during which time the Generator gets better and better at creating the fake data until it gets to a point where the fake data becomes almost indistinguishable from the real sample data.
"Data" as used in my explanation above could take many forms, one of the most common at this time being images.
In another blog post I will talk more about GANs and some of the more popular applications of the technology but today I just want to highlight how good GANs have become in generating fake data as can be seen from the six images at the start of this article.
GANs were first used to generate fake images in 2014 but the output, though revolutionary at the time, was of low quality.
However, as evident in the images below, the technology has improved in leaps and bounds over the years and today the generated images can be in high definition and so lifelike that it borders on the uncanny and downright spooky. If you want to learn more about GANs, read this article. and this one.
And this is only just the tip of the iceberg.
Can you spot the fake face (or faces) among the six images below? By fake I mean generated (in other words, the person does not exist and has never existed).
If your answer was one (or more) out of A, B, C, D, E or F then you are right. Actually ALL the faces are fake and the people in the images do not exist, have never existed and most likely never will.
The faces were "imagined" and created (in high definition) by an Artificial Intelligence algorithm. Specifically it was created by a type of Deep Learning algorithm/model known as a GAN which is short for Generative Adversarial Network.
What is a GAN?
In concept, a GAN is a combination of two separate neural networks that compete against each other to ultimately produce "fake" data that is as close as possible to real sample data.While one of the networks (the Generator) generates the fake data, the other network (the Discriminator) tries to detect whether the generated data is fake or not.
If it succeeds in detecting correctly, the Generator tries again using the feedback from the Discriminator to improve the "quality" of the generated data.
This back and forth process goes on for several cycles during which time the Generator gets better and better at creating the fake data until it gets to a point where the fake data becomes almost indistinguishable from the real sample data.
"Data" as used in my explanation above could take many forms, one of the most common at this time being images.
In another blog post I will talk more about GANs and some of the more popular applications of the technology but today I just want to highlight how good GANs have become in generating fake data as can be seen from the six images at the start of this article.
GANs were first used to generate fake images in 2014 but the output, though revolutionary at the time, was of low quality.
However, as evident in the images below, the technology has improved in leaps and bounds over the years and today the generated images can be in high definition and so lifelike that it borders on the uncanny and downright spooky. If you want to learn more about GANs, read this article. and this one.
Application of GANs
Though still a new and developing technology, GANs have shown a lot of promise in many fields one of which is in the medical field where it has been used to generate fake images that are used in training AI systems used to detect disease like cancers from mammograms and reading of x-rays and ct-scans.And this is only just the tip of the iceberg.
Comments
Post a Comment