AI-EDI-SPACE
收藏资源简介:
AI-EDI-SPACE数据集由联合国教科文组织景观城市化教席和蒙特利尔大学共同创建,旨在通过街景图像评估公共空间的质量。该数据集包含7,833张来自蒙特利尔地区的街景图像,通过两阶段抽样策略确保图像的代表性。数据集的创建过程结合了公平、多样性和包容性(EDI)原则,通过与不同社区组织的合作,确保了数据的多视角和高质量。该数据集主要应用于公共空间质量评估领域,旨在解决现有数据集在多样性和公平性方面的不足。
The AI-EDI-SPACE dataset was co-developed by the UNESCO Chair in Landscape Urbanization and the University of Montreal, aiming to assess the quality of public spaces using street-view imagery. It contains 7,833 street-view images sourced from the Montreal region, with a two-stage sampling strategy adopted to ensure the representativeness of the images. The dataset's development process integrates the principles of Equity, Diversity, and Inclusion (EDI), and through collaboration with various community organizations, it guarantees multi-perspective coverage and high data quality. This dataset is mainly applied in the field of public space quality assessment, aiming to address the shortcomings of existing datasets in terms of diversity and equity.




