Boosting Architectural Generation via Prompts: Report
Xin Zhang, Wenwen Liu
TL;DR
The paper addresses how prompts guide AI-driven architectural image generation and proposes a vocabulary taxonomy to structure prompt design. It presents categories from perspectives, lighting, rendering, to materials, landscapes, and architects, arguing that deliberate prompt design improves quality, realism, and creativity of generated designs. The approach emphasizes designer expertise and iterative refinement to align outputs with design intent and client requirements. The work aims to empower designers to harness AI tools by providing concrete vocabularies and prompts for more innovative and targeted architectural imagery.
Abstract
In the realm of AI architectural design, the importance of prompts is becoming increasingly prominent. With advancements in artificial intelligence and large-scale model technology, more design tasks are being delegated to machine learning algorithms. This necessitates a method for designers to guide algorithms in producing their desired designs. Prompts serve as a guiding and motivational mechanism, playing a crucial role in AI-generated architectural design. This paper categorizes and summarizes common vocabulary used in architectural design, discussing how to craft effective prompts and their impact on the quality and creativity of generated results. Through careful prompt design, designers can better control the generated architectural design images, thereby achieving designs that are more aligned with requirements and innovative.