GitHub - brianrtaylor/cvos: Causal Video Object Segmentation

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 57 Commits
3rdparty		3rdparty
cpp		cpp
cvos_aid		cvos_aid
evaluation		evaluation
external		external
generic		generic
mex		mex
util		util
LICENSE		LICENSE
README		README
cvos.m		cvos.m
cvos_run.m		cvos_run.m
demo.m		demo.m
mex_clean.m		mex_clean.m
mex_setup.m		mex_setup.m
setup.m		setup.m

Repository files navigation

==============================================================================
1. INSTALLATION
==============================================================================
Dependencies:
    * VLfeat (NOT INCLUDED)
        (http://www.vlfeat.org/) 

    * FastEdgeDetection (INCLUDED)
        (http://research.microsoft.com/en-us/downloads/389109f6-b4e8-404c-84bf-239f7cbf4e3d/)
        "Structured Forests for Fast Edge Detection", P. Dollár and C. 
        Zitnick, ICCV13
    
    * Piotr's Computer Vision MATLAB Toolbox (INCLUDED)
        (http://vision.ucsd.edu/~pdollar/toolbox/doc/index.html)
    
    * Accurate fast marching toolbox (INCLUDED)
        (http://www.mathworks.com/matlabcentral/fileexchange/24531-accurate-fast-marching)
    
    * Graph Based Image Segmentation (INCLUDED)
        (http://cs.brown.edu/~pff/segment/)
        "Efficient Graph-Based Image Segmentation" P. Felzenszwalb, D. 
        Huttenlocher, IJCV04
        Their package is C++ only. We include a MATLAB wrapper.

    * Recursive Bilateral Filtering (INCLUDED)
        ( http://www.cs.cityu.edu.hk/~qiyang/publications/eccv-12/ )
        "Recursive Bilateral Filtering", Q. Yang, ECCV12
        Their code does not allow the input to have only two channels (like
        flow does), so we modified their version to filter two-channel images 
        (as a special case), and also included a MATLAB wrapper.

    * SC - powerful image rendering for visualizing output (INCLUDED)
        (http://www.mathworks.com/matlabcentral/fileexchange/16233-sc-powerful-image-rendering)

You must run "mex_setup.m" in prior to anything

==============================================================================
2. FINAL WORDS
==============================================================================

This code was tested on Ubuntu.
If you use this work in your research, we ask that you cite the following:

Causal Video Object Segmentation from Persistence of Occlusions
B. Taylor, V. Karasev, and S. Soatto. In CVPR, 2015.

Website available at: http://brianrtaylor.github.io/cvos/
Additionally, if you have any comments about this work, please do not 
hesitate to contact us.

[email protected], [email protected]