This package contains a set of about 30k affine correspondences
(10k positive correspondences and 20k negative correspondences) in
24 image pairs.

This set was used in learning statistical models for the algorithm
which verifies correspondences by means of pixel-wise
cosegmentation. The method isdescribed in: Jan Cech, Jiri Matas,
Michal Perdoch; Efficient Sequential Correspondence Selection by
Cosegmentation; In CVPR, 2008, and is also available from
http://cmp.felk.cvut.cz/~cechj

@InProceedings{Cech-CVPR-2008,
 author = {Jan {\v C}ech, Ji{\v r}{\' i} Matas, Michal Per{\v d}och},
 title = {Efficient Sequential Correspondence Selection by Cosegmentation},
 booktitle = {IEEE Computer Society Conference on Computer Vision and Pattern Recognition},
 year = 2008,
 month = {June},
 venue = {Anchorage, Alaska, USA},
 keywords = {correspondence, verification, dense stereo, growing, sequential decison, learning, SVM, wide-baseline stereo, image retrieval, SIFT, RANSAC},
 psurl = {<a href="ftp://cmp.felk.cvut.cz/pub/cmp/articles/cech/Cech-CVPR-2008.pdf">[PDF]</a> },
 ISBN = {978-1-4244-2243-2},
 publisher = {Omnipress},
 address =  {Madison, WI, USA},
}

Please refer to that article if you are publishing results based
on this data.

Run demo_corresp to visualize the affine correspondences and to
see the format of the dataset.

This is free dataset for a non-commercial purpose. Commercial use
requires a licence. The dataset is distributed AS IS, WITHOUT ANY
IMPLIED OR EXPRESSED WARRANTY and WITHOUT ANY FURTHER SUPPORT for
the desired and/or other purpose.

(c) Jan Cech (cechj@cmp.felk.cvut.cz) FEE CTU Prague, Apr 30, 2009
