Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-scale Intervention Planning based on Generative Design

Ioannis Kavouras, Ioannis Rallis, Emmanuel Sardis, Eftychios Protopapadakis, Anastasios Doulamis, Nikolaos Doulamis

TL;DR

Urban green space scarcity in cities motivates new planning approaches. The authors apply generative AI, using image-to-image and image inpainting, to visualize multi-scale nature-based interventions on urban alleys via a GUI-based workflow. Results show that image-to-image generates photorealistic scenes aligned with prompts but can alter scale, while image inpainting preserves the intervention area and tends toward more implementable options, albeit with occasional architectural incompleteness. The work demonstrates a rapid ideation tool for architects and urban planners, while highlighting the need for domain-specific data and further refinement to improve real-world applicability.

Abstract

The scarcity of green spaces, in urban environments, consists a critical challenge. There are multiple adverse effects, impacting the health and well-being of the citizens. Small scale interventions, e.g. pocket parks, is a viable solution, but comes with multiple constraints, involving the design and implementation over a specific area. In this study, we harness the capabilities of generative AI for multi-scale intervention planning, focusing on nature based solutions. By leveraging image-to-image and image inpainting algorithms, we propose a methodology to address the green space deficit in urban areas. Focusing on two alleys in Thessaloniki, where greenery is lacking, we demonstrate the efficacy of our approach in visualizing NBS interventions. Our findings underscore the transformative potential of emerging technologies in shaping the future of urban intervention planning processes.

Multi-scale Intervention Planning based on Generative Design

TL;DR

Urban green space scarcity in cities motivates new planning approaches. The authors apply generative AI, using image-to-image and image inpainting, to visualize multi-scale nature-based interventions on urban alleys via a GUI-based workflow. Results show that image-to-image generates photorealistic scenes aligned with prompts but can alter scale, while image inpainting preserves the intervention area and tends toward more implementable options, albeit with occasional architectural incompleteness. The work demonstrates a rapid ideation tool for architects and urban planners, while highlighting the need for domain-specific data and further refinement to improve real-world applicability.

Abstract

The scarcity of green spaces, in urban environments, consists a critical challenge. There are multiple adverse effects, impacting the health and well-being of the citizens. Small scale interventions, e.g. pocket parks, is a viable solution, but comes with multiple constraints, involving the design and implementation over a specific area. In this study, we harness the capabilities of generative AI for multi-scale intervention planning, focusing on nature based solutions. By leveraging image-to-image and image inpainting algorithms, we propose a methodology to address the green space deficit in urban areas. Focusing on two alleys in Thessaloniki, where greenery is lacking, we demonstrate the efficacy of our approach in visualizing NBS interventions. Our findings underscore the transformative potential of emerging technologies in shaping the future of urban intervention planning processes.
Paper Structure (6 sections, 5 figures, 1 table)

This paper contains 6 sections, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related work
  3. Experimental Setup
  4. Experimental Results
  5. Conclusions
  6. Acknowledgments.

Figures (5)

  • Figure 1: A brief overview of the Image to Image workflow. Input: (a) Base Image; and (b) Description Text Prompt. Output: N-Generated Images indicating different scenarios of intervention planning.
  • Figure 2: A brief overview of the Inpainting Image workflow. Input: (a) Base Image; (b) Image Mask; and (c) Description Text Prompt. Output: N-Generated Images indicating different scenarios of intervention planning.
  • Figure 3: The images used for the experiments. The area secluded by the yellow dashed line corresponds to the masked area for the image inpainting technique.
  • Figure 4: The generated results for the Case Study 1. (a) Image-to-Image results; and (b) Image Inpainting results
  • Figure 5: The generated results for the Case Study 2. (a) Image-to-Image results; and (b) Image Inpainting results