Integrating Shape Grammar and Transformer Architectures for Automated Text-to-BRep Conversion in Design Automation

This project aims to create a system that can turn simple text descriptions of designs into detailed 3D models automatically. By combining shape grammar rules (which act like guidelines for building shapes) with advanced language-processing AI models called transformers, the system will understand natural language inputs and generate precise 3D representations known as Boundary Representation (Brep) models. For example, if someone describes a “three-story building with large windows and a flat roof,” the system will produce an accurate 3D model of that building. This innovation will make it easier for designers and architects to bring their ideas to life quickly and accurately. The participating institutions will benefit by advancing research in artificial intelligence and design automation, fostering collaboration between experts in computational design and AI, and potentially developing new tools that can be used in industry and education to streamline the design process.

Faculty Supervisor:

Yong Zeng

Student:

Partner:

Georgia Institute of Technology

Discipline:

Computer science

Sector:

Artificial Intelligence; Information and Communications Technology

University:

Concordia University

Program: