Apple just revealed a new artificial intelligence model called LiTo that can take a standard flat photograph and turn it into a fully detailed three-dimensional object. What makes this tool completely different from older scanning methods is how it handles light. Instead of just wrapping a flat picture around a digital block, LiTo actually understands how real light hits different materials, perfectly recreating natural shadows and bright reflections from just one single viewing angle.
How does the Apple AI understand physical materials and light
Before this development, creating a digital object required taking dozens of photos from every possible angle. Even then, the resulting digital model often looked weird when dropped into a new virtual environment because the original lighting was baked directly into the texture. LiTo solves this by separating the object itself from the light hitting it. The software looks at the original image and guesses the physical shape and the surface material at the same time.
If the picture shows a shiny metal coffee cup, LiTo knows how that specific metal should reflect light in any virtual room, whether it is sitting under a bright studio lamp or a dim street light.
What this new visual technology means for developers worldwide
Tech companies around the world have spent years trying to make augmented reality feel more natural. Apple is clearly positioning LiTo to help build out the global software ecosystem for spatial computing devices like the Vision Pro. Game developers and online retailers across different countries will soon be able to build massive libraries of digital items without needing expensive hardware.
An independent furniture maker could snap one picture of a new chair with a phone, and the software would instantly generate a perfect digital copy. Customers anywhere can then drop that digital chair into their living rooms to see how it catches the sunlight from their own windows.
Apple has not announced exactly when developers can download LiTo for consumer apps, but the research shows a massive leap forward in making digital items look real on a global scale.
