Remember in the Halloween special of “The Simpsons” when Homer walked through a portal and instantly became three-dimensional? Well, Stability AI has introduced a new tool that uses AI to do just that.
The new Stable Fast 3D AI model can produce a 3D asset from a single image in about half a second, which is a staggering speed compared to typical rendering systems. And even if it takes more than a second and a single image to perfect the 3D model, there is huge potential demand for these kinds of tools in everything from architecture to video games.
Stable Fast 3D produces a 3D model of an image that is not just a digital representation. The model’s high-quality UV unwrapped mesh ensures that textures are accurately mapped to 3D models, while reduced lighting tangling in textures results in more realistic and visually appealing assets. The result also includes options for colors and the material that covers the object, as soft, pale fur looks very different from a coating of aluminum foil.
From photos to movies
This isn’t Stability AI’s first 3D mode creator, but it is significantly faster than the earlier SV3D model, which took up to 10 minutes. Stable Fast 3D could theoretically create nearly 1,200 models and variants in 10 minutes. But the company can’t rest on its laurels, as other generative AI developers are experimenting with similar ideas. OpenAI’s POINT-E model can create 3D models from text prompts, and Adobe has its own experiments that can be implemented in its Firefly service.
It’s easy to see why there might be interest in 3D asset makers at this level and speed. The reduction in time, and therefore money, in creating these assets would appeal to many designers. It could help produce digital environments for virtual reality and games, enable experimentation in building design and room decoration, and even give boutique toy makers the chance to play with ideas that can be translated into the real world without the expense of producing a bunch of failed prototypes. Even online shopping could change, as you could potentially browse through interactive, 3D renderings of products.
You might also like…
- OpenAI unveils AI tool for building 3D models
- Generative AI could become more active thanks to this wild Stable Diffusion update
- Nvidia’s Latest AI Model Could Revolutionize Game Development