Creating detailed 3D models from 2D video

Not quite three months ago, Nvidia researchers published a paper focused on the use of image-based neural rendering for creating dense 3D surfaces. Their solution, called Neuralangelo, can recover dense 3D surface structures from multiview 2D images, with a high level of fidelity. As a result, it can generate detailed large-scale scene reconstructions from 2D video captures. In mid-August, Nvidia open-sourced Neuralangelo, releasing the code on GitHub. (Source: Nvidia) What do we think? The need for high-level computer-generated content is growing by leaps and bounds. Companies across all industries have adopted digital strategies, and the use of digital twins and