ModelNet40-C

该数据集对ModelNet40测试集中的对象应用了15种不同的污染方式，总计包含2,468个对象。该数据集被用于评估PointDR协议的鲁棒性，其规模达到2,468个对象，任务是对受污染的点云进行分类。This dataset applies 15 distinct corruption modalities to objects from the ModelNet40 test set, comprising a total of 2,468 objects. It is utilized to evaluate the robustness of the PointDR protocol, with the core task of classifying corrupted point clouds.