We Are Truly Fucked: Everyone Is Making AI-Generated Fake Porn Now
A redditor named 'deepfakes' quietly enjoying his hobby: Face-swapping celebrity faces onto porn performers’ bodies. He made several convincing porn videos of celebrities—including Gal Gadot, Maisie Williams, and Taylor Swift—using a machine learning algorithm, his home computer, publicly available videos, and some spare time.
Since last year, the practice of producing AI-assisted fake porn has exploded. More people are creating fake celebrity porn using machine learning, and the results have become increasingly convincing. Another redditor even created an app specifically designed to allow users without a computer science background to create AI-assisted fake porn. All the tools one needs to make these videos are free, readily available, and accompanied with instructions that walk novices through the process. The subreddit group, that deepfakes named after himself, has already amassed more than 15,000 subscribers. Within the community, the word “deepfake” itself is now a noun for the kinds of neural-network generated fake videos their namesake pioneered.
Training
The software is based on multiple open-source libraries, like Keras with TensorFlow backend. To compile the celebrities’ faces, deepfakes said he used Google image search, stock photos, and YouTube videos. Deep learning consists of networks of interconnected nodes that autonomously run computations on input data. In this case, he trained the algorithm on porn videos and Gal Gadot’s face. After enough of this “training,” the nodes arrange themselves to complete a particular task, like convincingly manipulating video on the fly.
The ease with which someone could do this is frightening. Aside from the technical challenge, all someone would need is enough images of your face, and many of us are already creating sprawling databases of our own faces: People around the world uploaded 24 billion selfies to Google Photos in 2015-2016. It isn’t difficult to imagine an amateur programmer running their own algorithm to create a sex tape of someone they want to harass.
Another user, called 'deepfakeapp,' even created FakeApp, a user-friendly application that allows anyone to recreate these videos with their own datasets. The subreddit’s wiki states that FakeApp is “a community-developed desktop app to run the deepfakes algorithm without installing Python, Tensorflow, etc.,” and that all one needs to run it is a “good GPU [graphics processing unit, the kind that high-end 3D video games require] with CUDA support [NVIDIA’s parallel computing platform and programming model].” If users don't have the proper GPU, they can also rent cloud GPUs through services like Google Cloud Platform. Running the entire process, from data extraction to frame-by-frame conversion of one face onto another, would take about eight to 12 hours if done correctly. Other people have reported spending much longer, sometimes with disastrous results.
- source: motherboard.vice.com
Other stuff:
– Amazing girl from Lithuania! Check her photos!
– Binance is opened for registration again!
– Would Ripple become 1000 USD coin?
– How to buy bitcoins? Complete guide!
Things to write blogs about ;)