The team led by Fei-Fei Li at Stanford has released a technology called ZeroNVS, which achieves scene-level full-scene generation. They introduced a method known as a 3D-aware diffusion model for generating natural and diverse images. By evaluating with metrics such as LPIPS, they have achieved remarkable results in zero-shot performance. This technological breakthrough signifies that AI can generate 360-degree full-scene images from a single image, which will have a profound impact in fields such as virtual reality and game development.