Motion segmentation data set for moving cameras
We present a new data set of moving camera videos for evaluating motion segmentation algorithms.
The zip file contains 5 videos taken with a moving camera in challenging
background scenarios.
Human labeled ground truth is provided for
evaluation of foreground/background classification accuracy.
If you use this data set, please cite the following paper:
Manjunath Narayana, Allen Hanson, Erik Learned-Miller.
Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations
International Conference on Computer Vision (ICCV), 2013.
[pdf]
Sample video output results for these and other benchmark videos
can be seen in the zipped supplementary material
[zip file].