Nvidia researchers have developed an artificial intelligence (AI) system that can turn 2D videos into highly detailed 3D graphics. With this tech, it can replicate structures like buildings, sculptures and other real-world artefacts. The sophisticated AI can create lifelike replicas of buildings, sculptures and other actual objects. It generates 3D structures with intricate textures and details that are so detailed that creative professionals can easily import them into design applications. They can then edit the objects for use in video game development, industrial digital twins, robotics, art and other create pursuits. The final 3D object can come in handy in creating virtual reality applications. Nvidia’s AI can also aid developers and other creatives to recreate real-world artefacts in the digital world. The tool will help them to improve their 3D reconstructions and enable them to import objects of different sizes and shapes into virtual environments.
Nvidia is a technology company that provides processors for gaming, artificial intelligence, design, robotics, and self-driving cars among other applications.
Ming-Yu Liu is a senior director of research and a co-author on the paper regarding the technology.
