* This dataset supports online use.Click here to jump.

Caltech-101 Dataset is a dataset consisting of 101 categories of object images, which is mainly used for target recognition and image classification. There are 40 to 800 images in different categories, each image is 300 * 200 pixels in size, and the publisher of the dataset has marked the corresponding targets for use.

The dataset was collected by Feifei Li, Marc Andreto and Marc'Aurelio Ranzato of California Institute of Technology in September 2003. Related papers include "Learning generative visual models from few training examples: an incremental Bayesian approach tested on 101 object categories" and "One-Shot learning of object categories".