Subscribe to updates

Just when you thought Autodesk couldn’t be any more all-in on AI, the company has doubled down and unveiled Project Bernini.

News that comes hot on the heels of the release of Autodesk’s free generative Fusion plugin, Project Bernini is a “research effort” for generative AI in the “Design and Make” industries. That effort’s first output is a generative AI model that creates 3D designs “from a variety of inputs including a single 2D image, multiple images showing different views of an object, point clouds, voxels, and text”.

The model is experimental only – and thus not available to the public – but it is nevertheless aimed at professional applications, with variation and real-world functionality the focus of the designs it creates. The given example is a water pitcher, which, with other AI models, would be generated as a shape with aesthetics-enhancing textures.

Project Bernini’s model instead separately generates texture and shape, resulting in a water pitcher that is actually hollow, and could theoretically be used, should it be physically created. The model also generates several variations of each design.

The model was trained on 10 million 3D shapes, comprised of “publicly available data, a mixture of CAD objects and organic shapes”. And it’s only the first stop for Project Bernini, Autodesk says.

How future models may differ is yet to be revealed, although “design, manufacturing, architecture, engineering, and construction … [and] media and entertainment” are all target industries. The project serves as one piece of a broader strategy under the “Autodesk AI” umbrella, which means we can probability expect a lot more AI-related news from the company.

For more on Project Bernini and its latest model, read Autodesk’s press release.

You’ve read that; now read this:

License: The text of "Autodesk and AI Meet Again With Project Bernini" by All3DP is licensed under a Creative Commons Attribution 4.0 International License.