arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.12564cs.CV
收藏

建筑中几何深度学习的合成3D数据生成流水线

Synthetic 3D Data Generation Pipeline for Geometric Deep Learning in Architecture

Politecnico di Milano(米兰理工大学);Computational Design Institute(计算设计研究所);IIIT Hyderabad(印度海得拉巴国际信息技术学院);KAUST(阿卜杜拉国王科技大学)

Stanislava Fedorova, Alberto Tono, Meher Shashwat Nigam, Jiayao Zhang, Amirhossein Ahmadnia, Cecilia Bolognesi, Dominik L. Michels

摘要:

随着建筑领域对深度学习算法和计算设计的兴趣日益浓厚,对大型、可访问且多样化的建筑数据集的需求也在增加。我们决定通过构建特定于现场的合成数据生成管道来解决这个问题,该管道生成任意数量的3D数据以及相关的2D和3D注释。注释的多样性、自定义生成的建筑物和数据集参数的灵活性使该框架适用于多个深度学习任务,包括需要直接3D监督的几何深度学习。在创建建筑数据生成管道时,我们利用了专家的架构知识,以构建一个模块化、可扩展的框架,并提供足够数量的类平衡数据样本。此外,我们有目的地让研究人员参与数据集定制,允许引入额外的建筑组件,材料纹理,建筑类,注释的数量和类型以及每个3D模型样本的视图数量。这样,该框架将满足不同的研究要求,并将适用于各种各样的任务。所有代码和数据都是公开的。

英文摘要:

With the growing interest in deep learning algorithms and computational design in the architectural field, the need for large, accessible and diverse architectural datasets increases. We decided to tackle this problem by constructing a field-specific synthetic data generation pipeline that generates an arbitrary amount of 3D data along with the associated 2D and 3D annotations. The variety of annotations, the flexibility to customize the generated building and dataset parameters make this framework suitable for multiple deep learning tasks, including geometric deep learning that requires direct 3D supervision. Creating our building data generation pipeline we leveraged architectural knowledge from experts in order to construct a framework that would be modular, extendable and would provide a sufficient amount of class-balanced data samples. Moreover, we purposefully involve the researcher in the dataset customization allowing the introduction of additional building components, material textures, building classes, number and type of annotations as well as the number of views per 3D model sample. In this way, the framework would satisfy different research requirements and would be adaptable to a large variety of tasks. All code and data are made publicly available.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑