Autodesk has unveiled Project Bernini, a research project technology that leverages AI to generate useful 3D models for professional BIM/CAD/3D workflows across industries.
The generative AI research aims to develop a technology that will allow professionals to generate in quick fashion very useful 3D models from various inputs, including text prompts, a single 2D image, or even multiple images of an object taken from different angles. Inputs could also include point clouds or voxels—which are 3D cubes located within an X-Y-Z three-dimensional coordinate system.
Project Bernini
Project Bernini is an effort from the Autodesk AI Lab within Autodesk Research. The company notes that a diverse set of scientists and specialists are working on finding ways for AI to tackle difficult and time-consuming tasks in manufacturing and AEC industries.
Project Bernini is generative AI research that enables sophisticated 3D model generation from text and image inputs into an AI engine.
The Bernini model has been trained on ten million 3D shapes made up of publicly available data. Part of the work was done in collaboration with the Chinese University of Hong Kong.
While not yet available for public use, Autodesk’s demonstration work in this arena is promising. To learn more, go here.