The OpenAI company, which specializes in artificial intelligence, has just developed a neural network that can generate photorealistic images. A simple sentence of text is enough to create almost any montage imaginable.
You will also be interested
[EN VIDÉO] Interview: How was artificial intelligence born? Artificial intelligence aims to mimic the functioning of the human brain, or at least its logic when it comes to making decisions. Jean-Claude Heudin, director of the research laboratory of the IIM (Internet and Multimedia Institute), explains the origin of this research.
OpenAI has just released a second version of its artificial intelligence dedicated to generating images. baptized DALL-E 2 (pronounced the painter Dalí), she is able to convert a simple text sentence into a photo-realistic image. The first version was content with a drawing on a plain background. This new AI creates much more complex compositions.
OpenAI is a direct competitor of Deep Mind from Google. This company dedicated to AI was founded by, among others, Elon Musk and investments received from Microsoft† His DALL-E 2 resembles the GauGAN (pronounced the painter Gauguin…) from Nvidia, being able to first convert a sketch into a photorealistic landscape and then do the same from a text sentence†
Explain how the DALL-E AI works with many examples. (Enable automatic subtitle translation in English.) © OpenAI
AI can also create variants of an existing image
However, DALL-E 2 is much more complex than the competition because it is not satisfied with landscapes. The AI can create an image that combines several common elements, such as ” a teddy bear skateboarding in Times Square “. The system is based on CLIP, a neural network from OpenAI trained on a large number of images with description. This AI is designed to analyze an image and come up with a description, but here it performs the opposite operation. A second step then decodes the result of the first to create a coherent image.
This two-stage system also offers other possibilities. The AI can take an existing image and replace an element, or create a variation inspired by the original by adjusting the angle, pose and aspect of the subject. However, to prevent abuse, the AI cannot generate photo-realistic images of human faces and the company has limited its ability to produce images with adult or violent content.
Interested in what you just read?