The recent wave of developments and research in the field of deep learning and artificial intelligence is causing the border between the intuitive and deterministic domains to be redrawn. Amidst all the excitement surrounding this field, there are several prototypes being made, most of which are narrow, single purpose applications of deep learning technologies. This thesis takes a step back to establish a broader understanding of the new class of algorithms that deep learning offers. Beginning with the observation that architectural design workflow is often characterized by several representational transformations as projects grow in resolution and complexity, from sketching to detailed drawings or models, this research developed a series of deep learning prototypes that illustrate the potential application of this technology in the larger design workflow. This paper discusses the performance of these prototypes, identifies the challenges for integrating deep learning in practical design applications. This paper also suggests some ways in which these technologies might affect how the design process is carried out.
展开▼
