Artificial intelligence is revolutionizing architecture through text-to-image synthesis, converting textual descriptions into detailed visual representations. We explore AI-assisted floor plan design, focusing on technical background, practical methods, and future directions. Using tools like, Stable Diffusion, AI leverages models such as Generative Adversarial Networks and Variational Autoencoders to generate complex and functional floorplans designs. We evaluates these AI models' effectiveness in generating residential floor plans from text prompts. Through experiments with reference images, text prompts, and sketches, we assess the strengths and limitations of current text-to-image technology in architectural visualization. Architects can use these AI tools to streamline design processes, create multiple design options, and enhance creativity and collaboration. We highlight AI's potential to drive smarter, more efficient floorplan design, contributing to ongoing discussions on AI integration in the design profession and its future impact.

翻译：人工智能正通过文本到图像合成技术革新建筑领域，将文本描述转化为精细的视觉呈现。本文探讨人工智能辅助的平面图设计，重点关注技术背景、实践方法及未来方向。借助Stable Diffusion等工具，人工智能利用生成对抗网络和变分自编码器等模型，生成复杂且功能完备的平面图设计方案。我们评估了这些人工智能模型通过文本提示生成住宅平面图的有效性。通过参考图像、文本提示和草图等实验，我们评估了当前文本到图像技术在建筑可视化中的优势与局限。建筑师可运用这些人工智能工具优化设计流程、创建多样化设计方案，并提升创造力与协作效率。我们强调人工智能在推动更智能、更高效的平面图设计方面的潜力，为设计专业中人工智能融合及其未来影响的持续讨论提供见解。