This AI turns images into 3D videos!

This is a new level reached with Stable Video 3D when it comes to creating videos using artificial intelligence. This revolutionary model impresses.

Stability AI expands its AI production catalog with Stable 3D video (SV3D). This new model converts images into 3D videos lasting a few seconds. Note that SV3D only needs one input image for the conversation. The British start-up impresses once again.

Remember that last November, Stability AI already surprised Stable video streaming. This template can create a video from a prompt or input image.

The generated video lasts two to five seconds. In addition, its frame rate is customizable. Also note that SV3D is based on this previous genetic AI.

Demonstrate stability AI expertise

SV3D is not solely based on Stable Video Diffusion technology. The new model also incorporates elements of it Fixed Zero123.

This is another model from the London start-up, available since last December. This generative AI makes it possible to easily create 3D images.

However, SV3D and Stable Zero123 do not not the same mechanism. The former can be seen as an improvement over the latter, he explains Varun Jabbani.

The Stability AI researcher explains that SV3D uses “ ONE new view composition network “. This takes a single image as input to create new view images.

Additionally, in a recent research paper, he further clarifies how Stable Video 3D works.

The main advantage of Stable Video 3D

Recent work on 3D modeling by AI provide techniques for adjust 2D production models for compositing new views and 3D optimization.

However, these techniques have their limitations. This results in limit the performance of 3D object modeling with artificial intelligence.

Stable Video 3D has the ability to consistently create multiple view images of an object. This allows you to get relevant shots from any angle.

So the new genetic AI from Stability AI falls short no problem with perspective to create 3D videos. Which makes it a very effective 3D modeling tool.

Two versions of SV3D available

For commercial purposes, Stability AI offers its new model at 20 dollars a month. Users with a turnover of less than one million dollars can benefit from this formula.

For non-commercial use, Stable Video 3D is available for free download on Embraced face.

On the other hand, the British start-up has designed two versions of its new genetic artificial intelligence: SV3D_u And SV3D_p.

On the one hand, SV3D_u allows you to get orbital or circular projection video from a single input image. This version does not require a camera conditioning.

In terms of genetic artificial intelligence, camera setup refers to a technique that uses a additional entrance to guide the process of creating new images.

This additional entry is often a second image or some settings related to camera perspectives.

On the other hand, SV3D_p extends this capability by taking both single images and circular views.

    Share this article:


Our blog is powered by readers. When you purchase through links on our site, we may earn an affiliate commission.

Leave a Reply

Your email address will not be published. Required fields are marked *