CAL500
收藏arXiv2025-09-30 收录
下载链接：
http://calab1.ucsd.edu/~datasets/
下载链接
链接失效反馈官方服务：
资源简介：
该数据集涉及的是多标签分类任务。在这个任务中，我们需要对每个样本进行标注，每个样本可能同时属于多个类别，这要求我们的分类模型不仅要能够识别出样本的主要特征，还要能够处理样本之间的复杂关系，确保每个标签都能被准确预测。
This dataset is intended for multi-label classification tasks. In such tasks, each sample requires annotation and may belong to multiple categories simultaneously. This mandates that the classification model must not only identify the core features of the sample but also address the intricate relationships between samples, ensuring that every label can be accurately predicted.
