Form data as a resource in architectural analysis: an architectural distant reading of wooden churches from the Carpathian Mountain regions of Eastern Europe

ABSTRACT

Recent research into architectural form analysis using deep learning (DL) methods has shown potential to identify features from large collections of building data, shedding new light into formal aspects of our built environment. As these methods begin to enter architectural, urban, and policy design contexts, it becomes important to develop critical approaches to employing them. In this paper, we document and reflect upon our efforts to create a custom dataset of 3-D models of 331 wooden churches located within the Carpathian Mountains of Eastern Europe, and to use DL methods to explore this dataset with the goal of revealing unexpected formal traits and advancing architectural scholarship on this subject. While existing scholarship groups them into four distinct stylistic categories, our analysis reveals stylistic overlaps, previously undetected micro styles, and shared architectural features. We posit the resulting analyses as an example of an ‘architectural distant reading’ that enriches our understanding of this architectural typology through an unprecedentedly detailed portrait of its formal characteristics based on a large architectural dataset. Crucially, drawing on recent developments in critical data and algorithm studies, we show how the dataset construction and subsequent analyses, and their results, were shaped by slow, manual data curation processes, methodological constraints, subjective decisions, and engagements with archives, domain experts. We thus illustrate how DL techniques might be contextualized for architectural studies in relation to other modes of knowledge and labour, and offer a detailed case study of state-of-the-art computational methods enriching established approaches to architectural form and historical analysis.

KEYWORDS:

• Form analysis
• deep learning
• digital heritage
• architectural distant reading

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 A tripartite plan includes three linearly connected spaces: the narthex at the front, the nave in the middle, and the sanctuary at the rear.

2 NeRS is an algorithmic method that converts sparse images of objects into geometrically and texturally accurate water- tight reconstructions (Zhang et al. 2021).

3 3-D shape templates reflect the general width, length, and height of its target building and are used as a scale guide to improve 3-D reconstruction accuracy.

4 A voxel is a well-established volumetric representation for forms, and are binary-valued arrays on a regular 3-D grid having a value of 1 when it is occupied by the form, and 0 otherwise.

5 SDF assigns a float rather than binary integers to a voxel, expressing the surface information of polygon meshes by distance from an origin of the meshes in each voxel.

 

Additional information

Notes on contributors

Michael Hasey

Michael Hasey is a computational designer who works at the intersection of artificial intelligence and architectural design and recently completed a Master of Computational Design degree at Carnegie Mellon University. He holds a Bachelor of Architecture degree (HBAS) from the University of Waterloo and a Master of Architecture degree (M. ARCH) from McGill University.

Jinmo Rhee

Jinmo Rhee is a computational designer and architect interested in the integration of artificial intelligence and space design. Jinmo holds an MS in Computational Design from Carnegie Mellon University, and is currently pursuing the PhD in Computational Design. He received his Bachelor of Architecture from Korea National University of Arts (KNUA) and completed the Royal Institute of British Architects (RIBA) part I and II.

Daniel Cardoso Llach

Daniel Cardoso Llach is an Associate Professor in the School of Architecture at Carnegie Mellon University, where he chairs the Computational Design graduate programme and co-directs CodeLab. He holds a Bachelor of Architecture from Universidad de los Andes, Bogotá, and a PhD and MS (with honours) in Design and Computation from MIT. He has also been a research fellow at Leuphana (MECS), Germany, and a visiting scholar at the University of Cambridge, UK.