TR2016-078

Gaussian Conditional Random Field Network for Semantic Segmentation


    •  Vemulapalli, R., Tuzel, C.O., Liu, M.-Y., Chellappa, R., "Gaussian Conditional Random Field Network for Semantic Segmentation", IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016, pp. 3224-3233.
      BibTeX TR2016-078 PDF
      • @inproceedings{Vemulapalli2016jun,
      • author = {Vemulapalli, Raviteja and Tuzel, C. Oncel and Liu, Ming-Yu and Chellappa, Rama},
      • title = {Gaussian Conditional Random Field Network for Semantic Segmentation},
      • booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
      • year = 2016,
      • pages = {3224--3233},
      • month = jun,
      • url = {https://www.merl.com/publications/TR2016-078}
      • }
  • Research Areas:

    Artificial Intelligence, Computer Vision, Machine Learning

Abstract:

In contrast to the existing approaches that use discrete Conditional Random Field (CRF) models, we propose to use a Gaussian CRF model for the task of semantic segmentation. We propose a novel deep network, which we refer to as Gaussian Mean Field (GMF) network, whose layers perform mean field inference over a Gaussian CRF. The proposed GMF network has the desired property that each of its layers produces an output that is closer to the maximum a posteriori solution of the Gaussian CRF compared to its input. By combining the proposed GMF network with deep Convolutional Neural Networks (CNNs), we propose a new end-to-end trainable Gaussian conditional random field network. The proposed Gaussian CRF network is composed of three sub-networks: (i) a CNN-based unary network for generating unary potentials, (ii) a CNN-based pairwise network for generating pairwise potentials, and (iii) a GMF network for performing Gaussian CRF inference. When trained end-to-end in a discriminative fashion, and evaluated on the challenging PASCALVOC 2012 segmentation dataset, the proposed Gaussian CRF network outperforms various recent semantic segmentation approaches that combine CNNs with discrete CRF models.

 

  • Related News & Events