New technology uses neural networks to turn 2D video clips into detailed 3D structures for import into design applications; NVIDIA Research to demo at this year’s CVPR in Vancouver June 18-22.
Today, NVIDIA shared news of Neuralangelo, a new AI model from NVIDIA Research for 3D reconstruction using neural networks to turn 2D video clips into detailed 3D structures, generating lifelike virtual replicas of buildings, sculptures, and other real-world objects. The new model will be demoed at this year’s CVF Computer Vision and Pattern Recognition Conference (CVPR) in Vancouver June 18-22.
This new research builds on technology behind NVIDIA Instant NeRF, released last year and winner of SIGGRAPH’s Best Paper Award. Likened to Michelangelo sculptures created from blocks of marble, Neuralangelo generates 3D structures with intricate details and textures, allowing users to quickly create objects to import into design applications for use in art, video game development, robotics, and industrial digital twins.
“The 3D reconstruction capabilities Neuralangelo offers will be a huge benefit to creators, helping them recreate the real world in the digital world,” said Ming-Yu Liu, senior director of research and co-author on the paper. “This tool will eventually enable developers to import detailed objects — whether small statues or massive buildings — into virtual environments for video games or industrial digital twins.”
Watch "Digital Renaissance: Neuralangelo by NVIDIA Research Reconstructs 3D Scenes from 2D Video Clips:"
Check out the NVIDIA blog here for a more detailed look at Neuralangelo.