Variational Mean Field Inference for image segmentation.
This repository contains an implementation of the Variational Mean Field Inference algorithm for image segmentation. This implementation is utilising the contributions of the paper entitled "Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials" by Philipp Krahenbuhl et al.
Included is a CPU and GPU implementation of the algorithm utilising direct filtering for the message passing phase.
The code in this repository is distributed under a BSD license, except either where indicated otherwise or where the source does not contain a BSD license header.