Memory Efficient Max Flow

Related Publication

This code implements the MEMF algorithm described in the following paper

Memory Efficient Max Flow for Multi-label Submodular MRFs.
Thalaiyasingam Ajanthan, Richard Hartley, and Mathieu Salzmann.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016.
IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), March 2018.

Code Assumptions

1. The MRF energy has the following form
   $$E(x) = \sum \theta_{i}(x_i) + \sum \theta_{ij} (x_i, x_j)\ ,$$
   where $\theta_{ij} (x_i, x_j) = \gamma_{ij} \theta(|x_i - x_j|)$.
2. The code currently supports MRF with 4-connected grid structure only and nodes are labelled from 0 --> width * height - 1 in raw major ordering.

Contact

This code is for research purposes only, if you want to use it for commercial purpose please contact us.
Email: ajanthan {at} robots {dot} ox {dot} ac {dot} uk

Example Usage

To assist the user, example.cpp, Makefile and sample data files are provided. The following command runs the MEMF algorithm on 10x10 image with 5 labels, with quadratic pairwise potential.

memf.exe 10 10 5 <sample>/toy_unary_10_10_5.txt <sample>/toy_binary_4_10_10.txt <sample>/toy_binaryPot_10_10_5_l2.txt

Code Ocean Link