NVIDIA Corporation, the American Hardware and Software company, has recently announced the development of Neuralangelo, an AI tool that can turn 2D video clips into detailed 3D structures.
NVIDIA rose to prominence thanks primarily to its cutting-edge GPUs and motherboard chipset. But they have also been a major supplier of AI-powered technology. NVIDIA Research, the company’s research and experimentation team, has developed this new AI model.
The company held a demo to demonstrate the prowess of its new AI. The model requires a 2D video of any object from different angles. It then uses its neural rendering technology to reconstruct the object in 3D with great detail, accurately capturing repetitive texture patterns, homogenous colours, and minute colour variations. Users can further modify the 3D reconstructions and implement them into other projects.
Speaking about the potential use of this technology, Ming-Yu Liu, senior director of research at NVIDIA, said, “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.”
NVIDIA intends to present Neuralangelo and several other projects at the Conference on Computer Vision and Pattern Recognition (CVPR) on June 18-22 in Vancouver.