Source: Venture Beat
Video by Nvidia:
Nvidia Research has developed Neuralangelo, an AI model that transforms 2D videos into detailed 3D graphics structures. It generates lifelike virtual replicas of real-world objects, surpassing previous methods in fidelity. Creative professionals can import these objects into design applications for art, gaming, robotics, and more. Neuralangelo’s ability to capture complex material textures from 2D videos makes it valuable for rapid virtual object creation. The model will be presented at CVPR, along with other projects like DiffCollage, which generates large-scale content.