Caltech101

The dataset used in the paper is Caltech101, which is a natural image classification dataset. It contains 101 categories of natural images.

BibTex: