Towards an historically-informed automatic reconstruction of the architecture of Venice.
PhD project of Paul Guhennec.
The objective of this PhD thesis is to define a generic method for 4D reconstruction of the architecture of a city, based on an architectural knowledge system built by processing, aligning and automatically analyzing visual and textual sources. This thesis is conducted within the Doctoral School of Digital Humanities (EDDH) of EPFL and is part of the SNSF project “Parcels of Venice” which studies how to gather and redocument knowledge about the urban space and population of Venice over the past centuries.
Through inference, extrapolation, and conditional procedural modeling, the cosntitution of such formalized knowledge of documented parcels will provide the basis for historically-informed hypothetical reconstructions of parcels lying in informational darkness. Newly added information (e.g., from a new source documenting a parcel) will then be confronted with such hypotheses and provide the opportunity for a critical reading of the adopted semantic grammar.
This methodology, which aims to be open to any city with sufficient archival density, will be applied to the architecture of Venice – whose typological resilience and richness of textual and iconographic representations make it particularly suitable for this undertaking.
This work is carried out in the framework of the Doctoral School of Digital Humanities (EDDH) at EPFL and is part of the SNSF project “Parcels of Venice”, which investigates how to collect and re-document knowledge about the urban space and population of Venice over the past centuries.