Test image and sequence data

 

Description

   This site presents a number of both monocular and epipolar-rectified stereo images and sequences. This data is made freely available to computer vision researchers to test their favourite stereo or monocular algorithms.

 

  Stereo sequences are particularly important to test algorithms' temporal stability and robustness. This data could also be used to test image-based rendering, face-detection, tracking and/or background-subtraction techniques.

 

  Some sequences show "distracting" non-stationary background, other present difficult but realistic non-Lambertian effects and/or image-saturation. In a few cases ground-truth foreground/background segmentation is provided.
   Labelled Stereo Data    Description

     Stereo data with ground-truth segmentation from our CVPR05 award winning paper.

           IU.zip
 - sequence pair (avi), 320x240, 160 frames, 64 Mb, moving background

- ground-truth foreground/background segmentation provided
(every 5th frame)

          AC.zip
 - sequence pair (avi), 320x240, 200 frames, 63 Mb

- ground-truth foreground/background segmentation provided
(every 5th frame)

         IUJW.zip
 - sequence pair (avi), 320x240, 248 frames, 105 Mb, moving background

- ground-truth foreground/background segmentation provided
(every 10th frame)

          JM.zip
 - sequence pair (avi), 320x240, 470 frames, 165 Mb

- ground-truth foreground/background segmentation provided
(every 10th frame)

          VK.zip
 - sequence pair (avi), 320x240, 262 frames, 104 Mb

- ground-truth foreground/background segmentation provided
(every 10th frame)

          MS.zip
 - sequence pair (avi), 320x240, 325 frames, 114 Mb

- ground-truth foreground/background segmentation provided
(every 10th frame)
     Stereo data without ground-truth
            Michel.zip  - image pair (bmp),   320x240, 400Kb
            Michel1.zip  - image pair (bmp),   320x240, 400Kb
            IlkayJohn1.zip  - sequence pair (avi),   320x240, 160 frames, 105 Mb, moving background
            IlkayJohn2.zip  - sequence pair (avi),   320x240, 160 frames, 122 Mb
                  Antonio1.zip  - sequence pair (avi),   240x320, 340 frames, 137 Mb
           Antonio2.zip  - image pair (bmp),   240x320, 388Kb
           Jamie.zip  - sequence pair (avi),   320x240, 85 frames, 30 Mb
           Jamie1.zip  - sequence pair (avi),   320x240, 100 frames, 34 Mb
           Jamie2.zip  - sequence pair (avi),   320x240, 100 frames, 35 Mb
           Geoff.zip  - sequence pair (avi),   320x240, 48 frames, 19 Mb, moving background
           Simon.zip  - sequence pair (avi),   320x240, 45 frames, 17 Mb
            Phil.zip  - sequence pair (avi),   320x240, 61 frames, 25 Mb
           Grid.zip  - image pair (bmp),   640x480, 1.34 Mb
           Chairs.zip  - sequence pair (avi),  640x480, 150 frames, 170 Mb, lots of thin structures
           Book.zip  - image pair (bmp),   320x240, 413 Kb
           Book1.zip  - image pair (bmp),   506x380, 1.13 Mb
           Plant.zip  - image pair (bmp),   640x480, 1.34 Mb
       Toys.zip  - image pair (bmp),   506x140, 356 Kb, mixed pixels (fur)
   Labelled Monocular Data     Description

     Monocular sequences with ground-truth segmentation from our CVPR07 paper on Tree-based classification.

           TreeBasedClassData.zip
 - dozens of labelled monocular sequences

- ground-truth foreground/background segmentation provided for many but not all frames.
 

 

Scientific publications
  1. V. Kolmogorov, A. Criminisi, A. Blake, G. Cross, C. Rother. Bi-layer segmentation of binocular stereo video 2005 San Diego, CA, US Proc. IEEE Computer Vision and Pattern Recognition (CVPR).
  2. A. Criminisi, A. Blake. The SPS Algorithm: Patching Figural Continuity and Transparency by Split-Patch Search 2004 Washington DC, US Proc. IEEE Computer Vision and Pattern Recognition (CVPR).
  3. P. H.S. Torr, A. Criminisi. Dense Stereo Using Pivoted Dynamic Programming 2004 Image and Vision Computing (IVC).
  4. A. Criminisi, J. Shotton, A. Blake, C. Rother, P. H.S. Torr. Efficient Dense Stereo and Novel-view Synthesis for Gaze Manipulation in One-to-one Teleconferencing 2003 Cambridge, UK Microsoft Research.
  5. A. Criminisi, J. Shotton, A. Blake, P. H.S. Torr. Gaze Manipulation for One-to-one Teleconferencing Oct. 2003 Nice Proc. IEEE International Conference on Computer Vision (ICCV).
  6. A. Blake, P. H.S. Torr, I. Cox, A. Criminisi. Estimating uncertainty in dense stereo disparity maps Mar. 2003 Cambridge, UK Microsoft Research
back to Computer Vision @ MSRC

Web site designed and maintained by A. Criminisi