Publication
This dataset was created as part of the PhD project Machine Understanding of Architectural Space: From Analytical to Generative Applications at EPFL Media and Design Lab (LDM). It was derived from the Cubicasa5K dataset. The dataset contains preprocessed data and scripts to perform the transformation, sampling, and cleaning. This project explores machine learning techniques for understanding architectural spaces from isovist representation, ranging from analytical tasks such as spatial typification and semantic discovery to generative tasks such as spatial sequence synthesis.