Caltech-UCSD Birds 200

For a newer revision of this dataset with more images and annotations, see Caltech-UCSD Birds-200-2011

Details

Caltech-UCSD Birds 200 (CUB-200) is an image dataset with photos of 200 bird species (mostly North American). For detailed information about the dataset, please see the technical report linked below.

• Number of categories: 200

• Number of images: 6,033

• Annotations: Bounding Box, Rough Segmentation, Attributes

Some related datasets are Caltech-256, the Oxford Flower Dataset, and Animals with Attributes. More datasets are available at the Caltech Vision Dataset Archive.

Citation

If you use CUB-200 in your work, please cite the technical report:

• Welinder P., Branson S., Mita T., Wah C., Schroff F., Belongie S., Perona, P. “Caltech-UCSD Birds 200”. California Institute of Technology. CNS-TR-2010-001. 2010. download pdf

@techreport{WelinderEtal2010,
	Author = {P. Welinder and S. Branson and T. Mita and C. Wah and F. Schroff and S. Belongie and P. Perona},
	Institution = {California Institute of Technology},
	Number = {CNS-TR-2010-001},
	Title = {{Caltech-UCSD Birds 200}},
	Year = {2010}
}

Download

You can download the dataset using the links below:

• Images, 647 megabytes.

• Lists, with class names, train/test splits, etc.

• Annotations (segmentations and bounding boxes), 11 megabytes.

• Attribute Annotations (text format), 22 megabytes.

• Attribute Annotations (yaml format), 24 megabytes.

• README

Contact

Contact Peter Welinder about the dataset. Please use the email exactly as written below (replacing the AT with @):

• peter.welinder+CUB200 AT gmail.com

Publications

The following publications use the dataset. Please contact us if you are using the dataset, and we will add your paper to the list.

• Branson S., Wah C., Babenko B., Schroff F., Welinder P., Perona P., Belongie S., “Visual Recognition with Humans in the Loop”, European Conference on Computer Vision (ECCV), Heraklion, Crete, Sept., 2010. download pdf

• Welinder, P., Perona, P. Online crowdsourcing: rating annotators and obtaining cost-effective labels. Workshop on Advancing Computer Vision with Humans in the Loop at CVPR. 2010. download pdf