The topic of deepfakes isn’t new. We’ve been discussing this artificial intelligence technique to create fake images and videos for years. Deepfakes are becoming increasingly difficult to distinguish from real footage. Moreover, what’s worrying is that there are numerous tools available to create such content, making it accessible to almost anyone.
A recent viral phenomenon is Deep-Live-Cam, an open-source software that enables real-time creation of deepfakes for free. This tool produces convincing results and users can use it in live videos and other apps.
A Viral Tool for Creating Deepfakes
The tool, Deep-Live-Cam, is compatible with Windows and Mac computers. Its creators have combined various open source software packages, offering a user-friendly interface. To start, users need to load a photo of the face they want to use for the deepfake into the program.
The interface includes several buttons for enhancing results. After clicking Live, the inswapper and GFPGAN models will begin performing a real-time face swap. The program is designed to correct any defects produced by the process in real time, thereby improving the output result.
Deep-Live-Cam
The team behind the project suggests that Deep-Live-Cam can be integrated into programs, such as OBS, for live broadcasts with a substituted face. The creators also emphasize that they’ve shared the software package as “a productive contribution to the rapidly growing AI-generated media industry.”
According to the creators, the concept is to assist artists with tasks like animating a character or using the character as a model for clothing. They also assert that the model includes security restrictions preventing it from being used to create inappropriate content, such as images with nudity or images of war.
Risks and Challenges of Software for Creating Feepfakes
The Deep-Live-Cam software is available in a GitHub repository for those willing to go through a few installation steps. However, it’s not the only tool for creating deepfakes. Fal, for instance, is another free tool available.
Living in today’s world, we must be cautious about whether what we see is a real image or one generated by AI. It’s important to be aware of the existence of such tools and their potential for misuse, such as in scams or spreading disinformation.
It’s worth noting that some platforms are taking steps to clearly identify AI-generated content. You’ve probably observed content tags on Facebook, Instagram, and TikTok. Additionally, Google plans to ban apps that could be used to create deepfakes for fraudulent purposes.
This article was written by Javier Márquez and originally published in Spanish on Xataka.
Image | Xataka using Midjourney
Related | Deepfakes From Met Gala Prove We Have a Problem. Even Katy Perry’s Mom Thought They Were Real
View 0 comments