Beckman Institute Datasets for Computer Vision Research

  The Ponce Group

    Main Page
    Vision Research
    Robotics Research
    Publications
    Technical Reports
    Datasets
    Software

Fifteen Scene Categories

This is a dataset of fifteen natural scene categories that expands on the thirteen category dataset released by Fei-Fei Li. The two new categories are industrial and store. Classification results for the fifteen categories are presented in the following paper:
Svetlana Lazebnik, Cordelia Schmid, and Jean Ponce. Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2006, accepted.
Download (84 MB zip file)

Thanks to Fei-Fei Li and Aude Oliva for providing parts of this database.





3D Object Recognition Stereo Dataset

This dataset consists of 9 objects and 80 test images. The training images are stereo views for each of the 9 objects that are roughly equally spaced around the equatorial ring for each of them. The number of stereo views ranges from 7 to 12 for the different objects. The test images are monocular images of objects under varying amounts of clutter and occlusion and different lighting conditions.

Results on this dataset are reported in the following paper:
Akash Kushal and Jean Ponce. Modeling 3D Objects from Stereo Views and Recognizing them in Photographs. Proceedings of the European Conference on Computer Vision, 2006, to appear.
Browse and download





3D Photography Dataset

This is a collection of ten multiview data sets captured in our lab. Each set consists of 24 images of a single rigid object, together with camera parameters and extracted apparent contours for each image.

Reconstruction results for this dataset appear in the following paper:
Yasutaka Furukawa and Jean Ponce. Carved Visual Hulls for Image-Based Modeling. Proceedings of the European Conference on Computer Vision, 2006, to appear.
Browse and download





Visual Hull Datasets

This is a collection of visual hull datasets used in the following paper:
Svetlana Lazebnik, Yasutaka Furukawa, and Jean Ponce. Projective Visual Hulls. Submitted to International Journal of Computer Vision, 2006.
Browse and download





Birds

This database contains 600 images (100 samples each) of six different classes of birds. The images are color JPEG, of variable resolution. The classes (each in its own directory) are as follows:
  • Egret
  • Mandarin duck
  • Snowy owl
  • Puffin
  • Toucan
  • Wood duck
If you use this database in your own research, please cite the following paper:
Svetlana Lazebnik, Cordelia Schmid, and Jean Ponce. A Maximum Entropy Framework for Part-Based Texture and Object Recognition. Proceedings of the IEEE International Conference on Computer Vision, Beijing, China, October 2005, vol. 1, pp. 832-838.
Download:
birds.zip (43 MB)
File numbers for training, validation, and test images for each category.
File numbers of training pairs





Butterflies

This database contains 619 images of seven different classes of butterflies. The images are color JPEG, of variable resolution. The classes (each in its own directory) are as follows:
  • Admiral: 111 images
  • Black Swallowtail: 42 images
  • Machaon: 83 images
  • Monarch 1 (wings closed): 74 images
  • Monarch 2 (wings open): 84 images
  • Peacock: 134 images
  • Zebra: 91 images
If you use this database in your own research, please cite the following paper:
Svetlana Lazebnik, Cordelia Schmid, and Jean Ponce. Semi-Local Affine Parts for Object Recognition. Proceedings of the British Machine Vision Conference, September 2004, vol. 2, pp. 959-968.
Download:
butterflies.zip (43 MB)
File numbers of training pairs and validation images for each class







Comparative Evaluation

Object Recognition Database

This database features modeling shots of eight objects and 51 cluttered test shots containing multiple objects. The images are color JPEGs, the resolutions are 1.2 Mpix (1280 x 960) and 3.7 Mpix (2200 x 1700). If you use this database in your own research, please cite the following paper:
Fred Rothganger, Svetlana Lazebnik, Cordelia Schmid, and Jean Ponce. 3D Object Modeling and Recognition Using Local Affine-Invariant Image Descriptors and Multi-View Spatial Constraints. International Journal of Computer Vision, vol. 66, no. 3, March 2006, pp. 231-259.
When reporting your results on this data, please refer to the comparative evaluation of different state of the art recognition methods contained in this article.

Browse and download:
Image directory
Archive in .tgz format (108 MB)
The creation of this database was supported in part by the National Science Foundation under grants IIS-0308087 and IIS-0312438, the UIUC-CNRS Research Collaboration Agreement, the European FET-open project VIBES, the UIUC Campus Research Board, and the Beckman Institute. For questions, contact Fred Rothganger (rothgang -at- uiuc.edu).





Texture Database

The texture database features 25 texture classes, 40 samples each. All images are in grayscale JPG format, 640x480 pixels. If you use this database in your own research, please cite the following paper:
Svetlana Lazebnik, Cordelia Schmid, and Jean Ponce. A Sparse Texture Representation Using Local Affine Regions. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 8, pp. 1265-1278, August 2005.
Browse and download:
Example images (four per texture)
T01-T05.zip (45 MB)
T06-T10.zip (40 MB)
T11-T15.zip (44 MB)
T16-T20.zip (46 MB)
T21-T25.zip (39 MB)
The creation of this database was supported in part by the National Science Foundation under grant IIS-0308087, the European project LAVA, the UIUC-CNRS Research Collaboration Agreement, the UIUC Campus Research Board, and the Beckman Institute. For questions, contact Svetlana Lazebnik (slazebni -at- uiuc.edu).





Video Sequences

This dataset is used for research on Euclidean upgrades based on minimal assumptions about the camera (e.g. square pixels).

Browse and download:
Example reconstructions
Archive in *.tgz format (23 MB)





Links to External Datasets

Texture

Object Recognition

Image Libraries

Statistics of Natural Images

Machine Learning