DELAUNAY

DELAUNAY数据集由伯尔尼大学生理学系创建，包含53位知名艺术家的抽象艺术作品，总计11,503件。数据集通过整合多个博物馆和艺术机构的在线资源构建，旨在为心理物理学和机器学习研究提供一个介于自然图像和人工模式之间的中间数据集。该数据集的创建过程涉及从多个数据库中筛选和整理艺术作品，确保每件作品的唯一性和艺术家的标识。DELAUNAY数据集的应用领域包括研究人类和人工神经网络在视觉分类任务中的样本效率差异，以及探索艺术风格识别等。

The DELAUNAY Dataset was created by the Department of Physiology, University of Bern, and contains 11,503 abstract artworks from 53 renowned artists. Constructed by integrating online resources from multiple museums and art institutions, this dataset is designed to provide an intermediate dataset between natural images and artificial patterns for psychophysical and machine learning research. The development process of the DELAUNAY Dataset involved screening and curating artworks from multiple databases, ensuring the uniqueness of each piece and correct attribution of the corresponding artist. Application areas of the DELAUNAY Dataset include investigating differences in sample efficiency between human and artificial neural networks in visual classification tasks, as well as exploring art style recognition and other related research topics.