Nvidia’s Instant NeRF an instant success

The 3D technology named one of Time magazine’s Best Inventions of 2022.

Nvidia’s Instant NeRF allows users to create a 3D environment by inputting 2D images or videos. (Source: Nvidia)

Nvidia’s Instant NeRF was recently recognized by Time as one of the Best Inventions of 2022. The software was well received, earning over 10,000 stars on GitHub and has been downloaded tens of thousands of times since its release in August. The research paper on Nvidia’s neural network also won the Siggraph 2022 Technical Papers Award. What has everyone so excited about? The tool re-creates scenes in 3D in just a few minutes, speeding up what used to be an expensive and slow process.

Users can try it for free on GitHub.

Instant NeRF uses neural radiance fields, fully-connected neural networks that can take a user’s set of 2D images or video and turn them into 3D images or scenes. The output model uses the camera pose from the images to render the 3D space of a scene. It isn’t magic software—the output wholly depends on the input. A user can input more images or take high-quality photos with similar lighting and camera angles to limit the amount of work Instant NeRF needs to do.

“Before Nvidia Instant NeRF, creating 3D scenes required specialized equipment, expertise, and lots of time and money. Now it just takes a few photos and a few minutes.” The award-winning tool, free to use for noncommercial purposes, processes static images (or video files) into realistically rendered 3D scenes using a new AI-based technology called neural radiance fields, or NeRF.

One huge upside to Instant NeRF over photogrammetry is that Instant NeRF retains backgrounds. This emerging technology can render the bulk or skeleton of a scene and be polished and ready for use. Another plus is that the software lives up to its name—it is almost instant. If a user makes a mistake or needs to add more photos, they will know before committing hours of work.