Last Updated on
Look at this picture carefully! Do you recognize any of the faces? Perhaps a celebrity or a soccer player or cricketer or a character from Game of Thrones (GOT) – any guesses? – NO? Well I guess it is not your fault, as these persons have never been born. You heard it right; these are all imaginary faces of the persons who are not actual and do not exist in real life. Well the science is here.
Nvidia’s AI can Now Create ‘Photos’ of People who don’t Exist
NVIDIA, the immensely reputed graphics chip manufacturer have managed to develop an Artificial Intelligence (AI) that creates extremely detailed faces of people who don’t even exist in real life. This system is the cutting-edge example of how AI is blurring the line between the “real” and the “fabricated”.
In June 2017, an image generator used a fairly new type of algorithm called a Generative Adversarial Network (GAN) for its AI-created faces that could turn even the crudest doodle of a face into a more realistic looking image. Using the same algorithm of GAN, NVIDIA made use of CelebA-HQ’s photo database consisting of famous people to generate images of human faces that doesn’t exist in real life.
How does Generative Adversarial Network (GAN) Work?
Here two neural networks are essentially set in opposition against one another. One of the networks functions as a generative algorithm, while the other confronts the results of the first, playing an oppositional role.
For the case of NVIDIA CelebA-HQ system, initially, the generative network would create an image at a lower resolution. Then, the discriminator network would assess the work. As the system progressed, the programmers added new layers dealing with higher-resolution details until the GAN finally generated images of “unprecedented quality,” according to the NVIDIA team’s paper.
It is not a denying fact that the line between human and machine was, is and will continue to blur. Now, the only question is if it will eventually disappear altogether.
Top 5 Amazing Facts about AI and GAN Technology
- If the average person didn’t know the faces were AI based-generated, they could easily believe they belonged to living people. Well what were your views when you saw the picture first? Did you look for what I asked in the beginning of the article?
- Did you know that, a number of AIs use deep learning techniques to produce human-sounding speech? Google’s DeepMind has WaveNet, which can now copy human speech almost perfectly.
- Lyrebird’s algorithm is able to synthesize a human’s voice using just a minute of audio.
- AI robots now can supposedly understand and express human emotion. Examples of those include Hanson Robotics’ Sophiaand SoftBank’s Pepper.
- Smile Vector is a Twitter bot that can make any celebrity smile. It scrapes the web for pictures of faces, and then it morphs their expressions using a deep-learning-powered neural network.
Must Check this