One day, it will be possible to technologically extract 3D scenes from any video on YouTube where the object is shown from different sides, and then the boom of three-dimensional virtual worlds will begin. Everything that didn’t get caught on camera will be filled in by AI, trained essentially on similar videos. Technologically, we are already close – google 3D Gaussian Splatting for Real-Time Radiance Field Rendering
Hi, I'm Rauf Aliev. 