To be ready soon.

Frequently Asked Questions

To be ready soon.


  • NEW! [RAR package][ZIP package] The image-set cropped with the bounding boxes provided by the original dataset, and resized (aspect ratio preserved) so that the larger side is 300 pixels.
  • NEW! [RAR package][ZIP package] The ground-truth landmark locations on each image. We have re-compute the coordinates or landmarks so that they fit exactly the images we provide in the previous item.
  • NEW! [RAR package][ZIP package] The part segmentation on each image (pixel-wise map, 0 for backgroud, 1-15 for the parts indicated by the original dataset). They exactly fit the images we provide in the previous item.
  • NEW! [RAR package][ZIP package] The feature-set obtained with the best parameters mentioned in the paper. We have provided the RAW version of features (without normalization). You can use the Feature Normalization (FeaNorm) for better classification performance.
  • If you find any mistakes or bugs in the codes or datasets, please contact me.

Related Publications

  • Lingxi Xie, Qi Tian, Richang Hong, Shuicheng Yan and Bo Zhang, "Hierarchical Part Matching for Fine-Grained Visual Categorization", in IEEE International Conference on Computer Vision (ICCV), Poster (Acceptance Rate = 27.8%), pages 1641--1648, Sydney, Australia, 2013. [PDF] [Poster] [Spotlight Slide] [Slides] [BibTeX]


  • Lingxi Xie, Qi Tian and Bo Zhang, "Spatial Pooling of Heterogeneous Features for Image Applications", in ACM International Conference on Multimedia (ACM-MM), 2012.
  • Catherine Wah, Steve Branson, Peter Welinder, Pietro Perona, and Serge Belongie, "The Caltech-UCSD Birds-200-2011 Dataset", Technical Report, the University California, San Diego (UCSD), 2011.
  • Svetlana Lazebnik, Cordelia Schmid and Jean Ponce, "Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories", in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2006.