Is Generative AI a successor to human-in-the-loop perception and cognition experiments in urban design and planning?
Year: 2025
Type: article
Abstract: This paper focuses on Generative AI models, deemed capable of human-like contextual understanding and reasoning capabilities and their potential use as participants for urban perception and cognition exercises. In recent years, widely available urban datasets, larger respondent pools, and advanced AI solutions facilitated scaling such studies to larger extents. Emerging computational approaches, such as Generative AI, further hold significant promise in aiding researchers to explore urban environments and synthesize perceptual experiences. This communication provides a perspective on utilizing such methods in urban assessment studies, where human-in-the-loop assessment is required to gather assimilated thoughts in the form of ratings and descriptions.
Source: Journal of Urban Design
Author: Deepank Verma
Institution: Technische Universität Braunschweig
Language: English
Subfield: Urban Studies
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W4411133747
