Have you heard of Fei-Fei Li? She has made significant contributions to the AI field and is even known as the “Godmother of AI.” Recently, World Labs, the AI company she founded, announced that it has developed an AI system capable of generating 3D worlds from a single static image, claiming it can transform film, gaming, simulator production, and other digital physical world representations.
AI godmother Fei-Fei Li launches system that can generate 3D worlds from a single static image.
AI godmother Fei-Fei Li founded World Labs, an AI company focused on developing large world models. The company recently unveiled its first project: an AI tool that can turn a single static image into a 3D world. In the future, using this tool, you will be able to step into the 3D world of any image.
World Labs’ AI tool for 3D worlds can render instantly in the browser. You can experience the feeling of stepping into the 3D world of classic paintings in World Labs’ article. Since we couldn’t embed it, the editor used a screen recording to give everyone a tour of the 3D world of the image. The 3D world created by World Labs is 365 degrees, so even locations not shown in the image can be generated. Movement is done using the WASD keys, and you can look around by clicking and dragging with the mouse.
Currently, World Labs restricts the movement range. If you go beyond the allowed range, a blue halo will appear on screen and prevent you from moving forward.
World Labs says its 3D world AI tool has many advantages:
- Will not change arbitrarily: once the 3D world is generated, everything in it remains exactly as it is, and nothing changes even if the camera moves away from the scene.
- Real-time control: After the 3D scene is generated, you can move around in it in real time, and viewing any detail in the scene is no problem at all.
- Accurate geometry: The 3D worlds generated by World Labs adhere to basic 3D geometric physics rules, so nothing feels unreal.
World Labs’ 3D world AI tool can take you into paintings, generating the world within the artwork, letting you experience those time-honored works of art in a whole new way.
A few creators have already gotten early access to preview and use World Labs’ 3D world AI tool, testing its potential in generative AI workflows. For example, video creator Eric Solorio used the 3D world AI tool to easily position characters in 3D scenes and precisely control camera movement.
Others also carefully design camera paths through generated 3D worlds, combining tools like Midjourney, Runway, Suno, ElevenLabs, Blender, and CapCut with World Labs’ technology to complete their work.
If you’d also like to try out World Labs’ 3D world AI tool for yourself, click the link below to visit the World Labs official website and experience controlling the 3D world.
Go experience World Labs’ 3D world AI tool.
Source: KOCPC Chinese