NVIDIA's AI "Neuralangelo" Turns iPhone Videos Into Detailed 3D Models

NVIDIA's latest AI tool , Neuralangelo, can transform simple iPhone videos into surprisingly detailed 3D structures, objects and digital spaces. Taking inspiration from Michelangelo's famous sculptures, Neuralangelo uses AI to sculpt intricate 3D models from ordinary video footage.

The resulting 3D objects can then be imported into design software for applications in art, gaming, robotics and digital twins.

While similar to other Nvidia tools that convert photos into 3D, Neuralangelo pushes the boundaries by accurately capturing fine details, textures and patterns in much larger 3D structures- from video of sculptures to buildings.

It achieves this using "instant neural graphics", the same tech behind Nvidia's Impressive NeRF AI.

Even video from drones hovering over a large space can be transformed into a massive, highly detailed 3D object.

Neuralangelo analyzes several video frames capturing different angles, determines each frame's camera position, then creates an initial 3D representation. The model then optimizes and sharpens details before outputting the final 3D creation.

Nvidia believes Neuralangelo "will be a huge benefit to creators, helping them recreate the real world in the digital world." And creators agree, saying the ability to import such detailed 3D objects and spaces from simple videos will revolutionize gaming, industrial design and more.