The Local Intelligence project was presented at the 41st Education and Research in Computer Aided Architectural Design in Europe (eCAADe) conference, held in 2023 at TU Graz, Austria. Recognized as one of the most prominent international platforms in architectural computation, eCAADe brings together scholars, designers, and technologists exploring the intersection of digital technologies and the built environment.

Our presentation, titled "Demystifying the Patterns of Local Knowledge: The Implicit Relation of Local Music and Vernacular Architecture", introduced a speculative design framework in which multimodal AI systems are used to translate between regional music and architectural forms. This presentation focused not on the technical pipeline alone, but on the broader epistemological implications of using generative AI to interpret culturally embedded, tacit knowledge.

One of the core arguments of the presentation was that architectural culture—when viewed through the lens of rhythm, proportion, and local memory—can be expressed across multiple modalities, including music. By training a vision-based generative model using facade images and corresponding spectrograms from Rebetiko music, the project posed the question: Can AI identify and reinterpret the spatial logic embedded in sound?

We discussed how CycleGAN and Pix2Pix models were adapted to explore this question, presenting a set of cross-modal translations that blurred the boundaries between architectural imagery and musical structure. [Figure 3 in the publication visualizes a set of facade-to-spectrogram translations, showcasing how architectural form might be reimagined sonically.]

In the Q&A and informal discussions that followed the session, audience members raised critical points about the cultural validity of AI-based interpretations, the ethics of visual translation, and the speculative use of GANs for heritage-sensitive design. These conversations highlighted the interdisciplinary potential of the work, particularly within emerging conversations on AI as an interpretive actor in architecture.

By contributing to the broader eCAADe theme of Digital Design Futures, Local Intelligence offered an example of how computational tools might engage with cultural complexity—not by flattening difference, but by tracing it across mediums.

Publication

Başarır, L., Çiçek, S., & Koç, M. (2023). Demystifying the patterns of local knowledge: The implicit relation of local music and vernacular architecture. In Proceedings of the 41st Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe) (Vol. 1, pp. 791–800). Graz: TU Graz. pdf available here