Dipl.Ing. Matthias Standfest


PhD Research Summary


Matthias Standfest is an architect with main interests at the intersection of cultural studies and the architectural design process itself. Before researching and teaching at ETH Zurich, department Information Architecture he spent a year researching at FCL Singapore. Actually he is working on extracting high-level features from geometric input with large scale unsupervised learning, and therefore is applying both methods and theories known from automatic image labelling and unsupervised machine translations. Before working at ETH he did his diploma on multi agent systems in design according to an actor network theory point of view at TU Graz, studied Philosophy of Science and received a HTL diploma in mechanical engineering.

PhD Thesis: Unsupervised Symmetric Mapping of Urban Geometry

His current research is focusing on how to index urban geometries according to emergent high-level features with large scale unsupervised learning. Therefore he is working on the symmetric encoding of polygon mesh geometry. He is investigating how geometry can be analyzed with machine learning tools while it is still maintaining the ability of being rendered into concrete shapes. Thus he is researching how to ensure a seamless integration of data analysis into the design process of urban planners.

Teaching: Creative Data Mining