SFS-A68

SFS-A68数据集由维也纳技术大学设计计算组创建，包含68个数字3D公寓建筑空间布局模型生成的输入和地面实况图像。该数据集适用于开发深度学习模型进行空间功能分割，旨在自动化建筑信息模型（BIM）中的空间功能分类，减少人工输入错误和提高分析准确性。数据集创建过程中，通过BIM或CAD系统从多种来源选择并解释建筑平面图，确保数据质量。应用领域包括建筑安全、能源分析和可用面积分析等。

The SFS-A68 dataset was created by the Design Computing Group at Technische Universität Wien (TU Wien). It consists of 68 sets of input data and ground-truth images generated from digital 3D spatial layout models of apartment buildings. This dataset is suitable for developing deep learning models for spatial functional segmentation, with the objective of automating spatial function classification in Building Information Modeling (BIM), minimizing manual input errors, and enhancing analysis accuracy. During the dataset's creation, architectural floor plans were selected and interpreted from multiple sources via BIM or CAD systems to ensure data quality. Its application areas include building safety, energy analysis, usable area analysis, and so on.