room_ID: An architectonic image classifier tool correlating machine learning and the domestic space

room_ID: An architectonic image classifier tool correlating machine learning and the domestic space

Nogueira, Alex; Romão, Luís

Full Article:

There are still considerable gaps in the relationship between artificial intelligence (AI) and architecture, both in the field of computer-aided architectural design and the everyday spatial experience. Therefore, we sought to build a bridge through correlating machine learning (ML), one of the significant branches of AI, and the domestic space, probably the most experienced architectonic space. Thus, our utmost goal is to develop an architectonic image classifier tool that allows the computer to identify the rooms that generally compound an ordinary dwelling. Our approach includes a brief theoretical background, the development of the room_ID app, and, afterward, discussions are presented. In the stage of the room_ID application development, a five-phase framework is proposed: a) Class definition; b) Database criteria; c) Image collection; d) Tool development; and e) Preliminary validation. Therefore, the room_ID tool provides us with a possible way to recognize the specificities of architectonic spaces rendered as computational data.

There are still considerable gaps in the relationship between artificial intelligence (AI) and architecture, both in the field of computer-aided architectural design and the everyday spatial experience. Therefore, we sought to build a bridge through correlating machine learning (ML), one of the significant branches of AI, and the domestic space, probably the most experienced architectonic space. Thus, our utmost goal is to develop an architectonic image classifier tool that allows the computer to identify the rooms that generally compound an ordinary dwelling. Our approach includes a brief theoretical background, the development of the room_ID app, and, afterward, discussions are presented. In the stage of the room_ID application development, a five-phase framework is proposed: a) Class definition; b) Database criteria; c) Image collection; d) Tool development; and e) Preliminary validation. Therefore, the room_ID tool provides us with a possible way to recognize the specificities of architectonic spaces rendered as computational data.

Palavras-chave: - -

DOI: 10.5151/sigradi2023-417

Referências bibliográficas

• [1]

Como citar:

Nogueira, Alex; Romão, Luís; "room_ID: An architectonic image classifier tool correlating machine learning and the domestic space", p-1812-1822. In: . São Paulo: Blucher, 2024.
ISSN 23186968, DOI 10.5151/sigradi2023-417