Scholarlink

<p>This paper introduces a deep learning-based approach for automated subspace data extraction and matching in residential floor plans, aiming to streamline interior soft decoration design processes. Traditional methods, relying on manual CAD drawings, 3D models, and renderings, are time-consuming and often fail to align with customer preferences. To address this, we develop a system that leverages deep learning to identify floor plan subspaces, extract structural information, and match them with a pre-designed case database, providing customers with the most relevant design references. Our contributions include an automated subsystem for subspace identification and separation, a method for case data information extraction, a similarity recognition scheme between floor plan subspaces and cases, and a self-built dataset comprising residential floor plans and design cases. Experimental results demonstrate that our approach outperforms existing methods in semantic segmentation accuracy and design reference matching, significantly improving interior design efficiency, reducing iterative design adjustments, and enhancing customer satisfaction. The code and dataset can be obtained from <a href="https://github.com/zyegao/Automated-Subspace-Matching">https://github.com/zyegao/Automated-Subspace-Matching</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automated subspace matching for residential floor plans using deep learning: enhancing interior design efficiency

  • 期刊论文

  • 发表日期：

    2025年6月30日

查看全文
  • Zhongye Gao,
  • Furao Shen,
  • Jian Zhao

摘要

This paper introduces a deep learning-based approach for automated subspace data extraction and matching in residential floor plans, aiming to streamline interior soft decoration design processes. Traditional methods, relying on manual CAD drawings, 3D models, and renderings, are time-consuming and often fail to align with customer preferences. To address this, we develop a system that leverages deep learning to identify floor plan subspaces, extract structural information, and match them with a pre-designed case database, providing customers with the most relevant design references. Our contributions include an automated subsystem for subspace identification and separation, a method for case data information extraction, a similarity recognition scheme between floor plan subspaces and cases, and a self-built dataset comprising residential floor plans and design cases. Experimental results demonstrate that our approach outperforms existing methods in semantic segmentation accuracy and design reference matching, significantly improving interior design efficiency, reducing iterative design adjustments, and enhancing customer satisfaction. The code and dataset can be obtained from https://github.com/zyegao/Automated-Subspace-Matching.