Description:
Key Words:
Architectural Sampling,Deep Learning,3D,Discrete
Required Skills:
Intermediate Grasshopper, Intermediate Rhino, beginner Python
Required Software:
Rhino, Grasshopper
Required Hardware:
PC or Mac
Maximum number of participating students:
15
This workshop continues the experimentation from last year’s ‘Discrete Sampling’ workshop in its use of deep neural networks for an end-to-end learning and generating of 3-dimensional forms. Framed as a hybrid design workflow, we will leverage on the probabilistic relations of discrete architectural forms and the high-dimensional feature representations of deep generative models. The theory of ‘Architectural Sampling’, which underlies the workshop's computational design framework, will be introduced to provide the foundational conceptual understanding needed for the hands-on design project investigations.