Realistic videos of people can be made with just a photo thanks to Samsung's deepfake AI
Scandals involving heavily doctored photos of people is something commonly heard of nowadays. In the near future, it is likely that we’ll be seeing scandals and fake news involving videos thanks to Samsung’s deepfake AI development.
In Samsung’s AI lab located in Russia, the company’s AI had managed to create an extremely realistic fake video of people simply by using a single photo, image or painting. The video mentioned had used the Mona Lisa painting, putting her into a scenario where she talks and moves. As we all know, there is one painting with the Mona Lisa in it. The video below explains how the technology is able to produce this:
Samsung has dubbed this technology as “realistic neural talking heads”. The word neural refers to the neural network which is a machine that learns and mimics the human brain and shares something similar with Samsung’s new algorithm. The AI watches various videos in order to learn the movements and habits of human faces before applying what it knows to the photo in order to create the realistic fake video.
While it does create many opportunities such as better action movies, documentaries, research and more, this technology does create concern in terms of fake news and misinformation. Celebrities, public figures and important people’s photos can be manipulated not not only spread fake news but also to defame others which can be damaging to not only the person but also the society. What do you think? For the latest technological advancements, stay tuned to TechNave.com!