Neuralangelo by NVIDIA

Neuralangelo, a new AI model by NVIDIA Research, is designed for 3D reconstruction using neural networks. It turns 2D video clips into detailed 3D structures, making it ideal for creative professionals such as architects, game developers, and designers.
Neuralangelo works by selecting several frames from a 2D video of an object or scene filmed from various angles, then using instant neural graphics primitives to capture finer details. The model optimizes the render to sharpen the details, resulting in a 3D object or large-scale scene that can be used in virtual reality applications, digital twins, or robotics development.
Creative professionals, such as architects and game developers, get the most value from Neuralangelo as it enables them to rapidly create usable virtual objects for their projects using footage captured by smartphones. The high fidelity of Neuralangelo's 3D reconstructions makes it easier for developers to import detailed objects into virtual environments for video games or industrial digital twins.