Skip to main content

ThisPersonDoesNotExist has spawned a host of amazing copycat sites

Who can predict what’s going to evolve into a popular online meme? One day it might be a humorous celebrity reaction at an awards ceremony; the next it’s a cutting-edge Nvidia generative A.I. that’s capable of dreaming up original images out of nowhere. The second of these options sums up what happened recently when a new website titled popped up online. Using a pair of adversarial neural networks, it can generate highly convincing human faces which — you guessed it! — don’t actually exist.

Inspired by that website, a number of other generative adversarial network, or GAN, sites have now emerged, too, using the same StyleGAN machine learning tool to generate everything from slightly creepy cats to imaginary Airbnb listings to computer-created anime love interests. The GAN works by pitting neural networks against each other to generate new images, and then tries to spot which one is real and which is fake. Over time, the “generator” part of the GAN gets more adept at creating fakes.

“I was interested in ThisPersonDoesNotExist because I could not understand how such a thing could possibly work,” Christopher Schmidt, creator of ThisAirBNBDoesNotExist, told Digital Trends. “I figured that surely the model must be working from some template to make its images. Finding out that it was not — that it was truly able to take any random set of inputs, and convert it to a realistic-seeming representation of a face — was a huge surprise to me. I will completely admit I still don’t understand how it works, but I do understand that it works, and it’s an interesting space to be in.”

The technology behind all of these websites is the same, as is the approximate layout of the sites, which allow users to hit “refresh” as many times as they wish, and provides a 100-percent new image every time. However, as noted, the images created span a wide range of genres — and all come with their own challenges.

“I picked anime out of a mix of intrinsic amusement factor, availability of high-quality, data, and genuine difficulty for machine learning,” Gwern Branwen, creator of ThisWaifuDoesNotExist, told us. “Anime is not something you see in many published papers, and is something everyone can appreciate: bad results are funny, while good results are even funnier.”

If you’re interested in machine learning, the results of these myriad websites show just how impressive and versatile Nvidia’s GAN is. If you’re not, and just want to be entertained, then the oftentimes surreal results generated by imaginary cats, fake rental properties, and anime heartthrobs that no actual artist has created are nonetheless sure to hit the sweet spot.

Editors' Recommendations

Luke Dormehl
I'm a UK-based tech writer covering Cool Tech at Digital Trends. I've also written for Fast Company, Wired, the Guardian…
Deepfake-hunting A.I. could help strike back against the threat of fake news
Kim Kardashian Deepfake Interview Image

Sylvester Stallone deepfake (replacing Arnold Schwarzenegger in Terminator 2: Judgement Day) Ctrl Shift Face/Youtube

Of all the A.I. tools to have emerged in recent years, very few have generated as much concern as deepfakes. A combination of “deep learning” and “fake,” deepfake technology allows anyone to create images or videos in which photographic elements are convincingly superimposed onto other pictures. While some of the ways this tech has been showcased have been for entertainment (think superimposing Sylvester Stallone’s face onto Arnie’s body in Terminator 2), other use-cases have been more alarming. Deepfakes make possible everything from traumatizing and reputation-ruining “revenge porn” to misleading fake news.

Read more
Photorealistic A.I. tool can fill in gaps in images, including faces
pen net image infilling screen shot 2019 07 16 at 05 19 53

You only need to go check out the latest Hollywood blockbuster or pick up a new AAA game title to be reminded that computer graphics can be used to create some dazzling otherworldly images when called for. But some of the most impressive examples of machine-generated images aren’t necessarily alien landscapes or giant monsters, they’re image modifications that we don’t even notice.

That’s the case with a new A.I. demonstration created by computer scientists in China. A collaboration between Sun Yat-sen University in Guangzhou and Beijing’s Microsoft Research lab, they’ve developed a smart artificial intelligence which can be used to accurately fill in blank areas in an image: Whether that’s a missing face or the front of a building.

Read more
A.I. can spot galaxy clusters millions of light-years away
ai identify galaxy clusters 205077 web 1

Image showing the galaxy cluster Abell1689. The novel deep learning tool Deep-CEE has been developed to speed up the process of finding galaxy clusters such as this one, and takes inspiration in its approach from the pioneer of galaxy cluster finding, George Abell, who manually searched thousands of photographic plates in the 1950s. NASA/ESA

Galaxy clusters are enormous structures of hundreds or even thousands of galaxies which move together, and for many years they were some of the largest known structures in the universe (until superclusters were discovered). But despite their massive size, they can be hard to identify because they are so very far away from us.

Read more