Journal of Artificial Intelligence in Architecture
Vol. 5 No. 2 (2026): Artificial Intelligence for Architecture: From Sustainable Management to Digita

YOLO-Based Identification and AI Generative Design of Nusantara Architectural Element through Kisho Kurokawa’s Symbiosis Concept

David Ricardo (Department of Architecture, Institut Teknologi Sumatera, South Lampung, Lampung, Indonesia)
Guruh Kristiadi Kurniawan (Department of Architecture, Institut Teknologi Sumatera, South Lampung, Lampung, Indonesia)
Antusias Nuzukhrufa (Department of Architecture, Institut Teknologi Sumatera, South Lampung, Lampung, Indonesia)
Fasha Nurliansyah Mahendra (Department of Architecture, Institut Teknologi Sumatera, South Lampung, Lampung, Indonesia)



Article Info

Publish Date
18 Aug 2026

Abstract

This research is rooted in the fundamental practice of combining two forms, long pursued by the Japanese architect Kisho Kurokawa, who gave rise to the concept of Symbiosis (Combining two different elements). The purpose of this research is to utilise artificial intelligence, such as YOLO v8, for identification and diffusion models to produce new building images with the Nusantara and Kisho Kurokawa concepts. The problem of this research is how Kisho Kurokawa's concept can be integrated into Nusantara architecture by re-identifying its elements and transforming them. The purpose of this research is to identify and transform the facade with the Nusantara and Symbiosis concepts. The method used is experimental, involving the collection of Lampung building datasets from the Nusantara era, followed by initial identification using YOLO v8 and classification of character similarities based on the symbiosis concept. The final step is to transform the Nusantara form (stable diffusion) for re-identification with YOLO v8 against the Nusantara facade elements. The results show that YOLO v8 can still identify various models of Nusantara architectural elements, even when they have changed shape, based on the contextual similarity to Kisho Kurokawa's Symbiosis concept.

Copyrights © 2026






Journal Info

Abbrev

JARINA

Publisher

Subject

Civil Engineering, Building, Construction & Architecture Computer Science & IT

Description

Journal of Artificial Intelligence in Architecture (JARINA) is currently accepting manuscripts from professionals, teachers, researchers, and students in various backgrounds, including the following disciplines: Architecture Urban Design Building Sciences Informatics Engineering in Architecture ...