Artificial intelligence is gaining more and more ground, and despite the crass ignorance (in the matter) of certain Italian editorialists, the "little-known" Nvidia is a candidate to be one of the protagonists of this revolution. The latest news on the table? The announcement by Neuralangelo, an AI model capable of taking a 2D video and transforming it into a detailed 3D model.
How does Neuralangelo work?
Let's think about a designer who wants to make a 3D scan of a truck. Our friend will take his smartphone, take a video of the truck from different angles and... He will feed everything to Neuralangelo.
The AI software chooses the best frames and builds a 3D model of the truck, which can then be optimized and imported into a CAD program. And if we talk about more complex structures or interiors? No problem, just use videos shot by a drone and Neuralangelo does everything you need.
Small details, big differences
One of the aspects that make Neuralangelo unique is its attention to detail. According to Nvidia, in fact, the AI model is able to capture "fine" elements, such as glass panels, smooth marble or roof tiles, with surgical precision.
The level of detail achieved by the system is currently unattainable. Take a look for yourself at this formidable example of the “Digital Renaissance”, to paraphrase the title of this video.
The new era of AI
Neuralangelo isn't Nvidia's only ace up its sleeve: the tech giant is in extraordinary shape. Between new computing platforms and generative AI that produce realistic dialogues even for video game characters, you are spoiled for choice: and time is in your favor.
Thanks to generative models of AI like ChatGPT (powered by thousands of Nvidia graphics cards), the company saw its valuation rise to over $1 trillion, a milestone achieved for the first time by a chip company, joining Apple's exclusive club, Amazon and Microsoft.
A growth, it is appropriate to say, that Nvidia makes. Nooooo. Oh well, I've written it now.