Form Forge: Latent Space Exploration of Architectural Forms via Explicit Latent Variable Manipulation

TL;DR

Form Forge tackles the challenge of exploring high-dimensional latent spaces for architectural form design by enabling per-dimension latent control rather than projection-based navigation. It introduces a fine-tuned StyleGAN2-ADA model trained on Blanciak's SITELESS sketches and an interactive UI with $512$ latent-variable manipulators encircling the generated form. The work provides a complete pipeline—from dataset curation and visualization to model fine-tuning and an interactive front end with a saved-samples carousel and adjustable generation settings—demonstrating a tangible tool for creative exploration. It also discusses limitations related to latent disentanglement and outlines future improvements (super-resolution, 3D forms, diffusion/StyleGAN3, evaluation metrics) to enhance fidelity, interpretability, and user agency.

Abstract

This paper presents 'Form Forge,' a prototype of a creative system for interactively exploring the latent space of architectural forms, inspired by Franois Blanciak's SITELESS: 1001 Building Forms via direct manipulation of latent variables. Utilizing a fine-tuned StyleGAN2-ADA model, the system allows users to navigate an array of possible building forms derived from Blanciak's sketches. Distinct from common latent space exploration tools that often rely on projected navigation landmarks, Form Forge provides direct access to manipulate each latent variable, aiming to offer a more granular exploration of the model's capabilities. Form Forge's design is intended to simplify the interaction with a complex, high-dimensional space and to serve as a preliminary investigation into how such tools might support creative processes in architectural design.

Figures (2)

• Figure 1: Organized square sample images from SITELESS: 1001 Building Forms ( ?)
• Figure 2: Form Forge User Interface