Nvidia’s impressive new AI model can convert 2D snapshots to 3D scenes
In an impressive AI demo, Nvidia has introduced a tool that can convert 2D snapshots to 3D-rendered scenes. The tool is named Instant NeRF after the neural radiance fields technique.
This method involves mapping the color and light intensity of multiple 2D images to generate data that can connect the images and render a 3D scene. The system requirements include images and camera position information.
While such a technique is being researched for years now, Nvidia claims Instant NeRF to be the fastest. This almost instant conversion process can unlock a plethora of possibilities, such as training robots and self-driving cars. It can also benefit architecture and entertainment by generating digital renders of real environments for creators to work on.