Architectural Ambiance: ChatGPT Versus Human Perception DOI Open Access
Rachid Belaroussi, Jorge Martín‐Gutiérrez

Electronics, Год журнала: 2025, Номер 14(11), С. 2184 - 2184

Опубликована: Май 28, 2025

Architectural ambiance refers to the mood perceived in a built environment, assessed through human reactions virtual drawings of prospective spaces. This paper investigates use ready-made artificial intelligence model automate this task. Based on professional BIM models, videos tours typical urban areas were built: business district, strip mall, and residential area. GPT-4V was used assess aesthetic quality environment based keyframes characterize these spaces shaped by subjective attributes. The spatial qualities analyzed experience include space scale, enclosure, style, overall feelings. These factors with diverse set attributes, ranging from balance protection elegance, simplicity, or nostalgia. Human participants surveyed same questions videos. answers compared according Our findings indicate that, while demonstrates adequate proficiency interpreting spaces, there are significant differences between AI evaluators. In nine out twelve cases, AI’s assessments aligned majority voters. district proved more challenging assess, green effectively modeled.

Язык: Английский

Enhancing participatory planning with ChatGPT-assisted planning support systems: a hypothetical case study in Seoul DOI
Steven Jige Quan, Syng‐Ook Lee

International Journal of Urban Sciences, Год журнала: 2025, Номер unknown, С. 1 - 34

Опубликована: Фев. 14, 2025

Язык: Английский

Процитировано

1
Subjective Assessment of a Built Environment by ChatGPT, Gemini and Grok: Comparison with Architecture, Engineering and Construction Expert Perception DOI Creative Commons
Rachid Belaroussi

Big Data and Cognitive Computing, Год журнала: 2025, Номер 9(4), С. 100 - 100

Опубликована: Апрель 14, 2025

The emergence of Multimodal Large Language Models (MLLMs) has made methods artificial intelligence accessible to the general public in a conversational way. It offers tools for automated visual assessment quality built environment professionals urban planning without requiring specific technical knowledge on computing. We investigated capability MLLMs perceive environments based images and textual prompts. compared outputs several popular models—ChatGPT, Gemini Grok—to experts Architecture, Engineering Construction (AEC) context real estate construction project. Our analysis was subjective attributes proposed characterize various aspects environment. Four identities served as case studies, set virtual designed using professional 3D models. found that there can be an alignment between human AI evaluation some such space scale architectural style, more accordance with vegetation. However, were noticeable differences response patterns AIs AEC experts, particularly concerning emotional resonance identities. raises questions regarding hallucinations generative where invents information behaves creatively but its are not accurate.

Язык: Английский

Процитировано

0
Designing effective image-based surveys for urban visual perception DOI
Youlong Gu, Matías Quintana, Xiucheng Liang

и другие.

Landscape and Urban Planning, Год журнала: 2025, Номер 260, С. 105368 - 105368

Опубликована: Апрель 17, 2025

Язык: Английский

Процитировано

0
Architectural Ambiance: ChatGPT Versus Human Perception DOI Open Access
Rachid Belaroussi, Jorge Martín‐Gutiérrez

Electronics, Год журнала: 2025, Номер 14(11), С. 2184 - 2184

Опубликована: Май 28, 2025

Architectural ambiance refers to the mood perceived in a built environment, assessed through human reactions virtual drawings of prospective spaces. This paper investigates use ready-made artificial intelligence model automate this task. Based on professional BIM models, videos tours typical urban areas were built: business district, strip mall, and residential area. GPT-4V was used assess aesthetic quality environment based keyframes characterize these spaces shaped by subjective attributes. The spatial qualities analyzed experience include space scale, enclosure, style, overall feelings. These factors with diverse set attributes, ranging from balance protection elegance, simplicity, or nostalgia. Human participants surveyed same questions videos. answers compared according Our findings indicate that, while demonstrates adequate proficiency interpreting spaces, there are significant differences between AI evaluators. In nine out twelve cases, AI’s assessments aligned majority voters. district proved more challenging assess, green effectively modeled.

Язык: Английский

Процитировано

0