Skip to content
Lumiere, the artificial intelligence tool that creates realistic videos from text |  VIDEO

Lumiere, the artificial intelligence tool that creates realistic videos from text | VIDEO

Lumiere, the artificial intelligence tool that creates realistic videos from text |  VIDEO

The artificial intelligence (AI) continues to generate news. This time Lumiere has been discovered, a tool developed by Google that has the ability to create realistic videos from text or an image.

However, Lumiere is not a tool that is available to the public. This is a project that brings together different researchers from Google Research, Tel Aviv University, the Weizmann Institute and the Technion.

READ ALSO: MSI Titan GT, a gaming laptop with the latest generation Intel Core i9 processor: what new features does it present? | REVIEW

The results of this research can be seen on the GitHub portal, where you can also access the paper which shows the work around this artificial intelligence model.

The text points out that Lumiere “is designed to synthesize videos that portray realistic, diverse and coherent movements, a fundamental challenge in video synthesis.”

To achieve this, we have worked on a new architecture, different from what is known in AI video generation.

“We demonstrate state-of-the-art text-to-video generation results and show that our design facilitates a wide range of content creation tasks and video editing applications, including image-to-video conversion, video painting, and stylized generation,” points out the research.

READ ALSO: José, María, John…: these are the most used names in the world, according to artificial intelligence

What can this model do?

Lumiere would fulfill some functions such as stylized generation, based on a realistic reference image. Video stylization is also mentioned, with which you can request that the object or person that appears in it have a different style.

Something similar happens with the animation of already existing images. For example, Lumiere can animate a photo of a butterfly on a plant or a fire in a bonfire.

Other possibilities are the generation of real videos, with the intention of completing unfinished videos or with cut areas, or the modification in real time giving a description to the AI.

Source: Elcomercio

Share this article:
globalhappenings news.jpg
most popular