Digital-Dermatology/CleanPatrick

CleanPatrick是一个基于Fitzpatrick17k皮肤病学数据集构建的图像领域数据清洗基准数据集。它包含17,000个皮肤病学图像，由933名医学众包工作者提供了超过500,000个关于三种数据质量问题的二元标签：离题样本、近似重复和标签错误。数据集旨在为图像数据集，特别是医学领域的图像数据集提供现实世界的测试床，以基准数据清洗策略。

CleanPatrick is a benchmark dataset for data cleaning in the image domain, built on the Fitzpatrick17k dermatology dataset. It consists of 17,000 dermatology images annotated with over 500,000 binary labels across three data quality issues: off-topic samples, near-duplicates, and label errors, provided by 933 medical crowd workers. The dataset aims to provide a realistic test bed for benchmarking data cleaning strategies for image datasets, especially in the medical field.