Acceleration Strategies in Generalized Belief Propagation

Generalized belief propagation is a popular algorithm to perform inference on large-scale Markov random fields (MRFs) networks. This paper proposes the method of accelerated generalized belief propagation with three strategies to reduce the computational effort. First, a min-sum messaging scheme and a caching technique are used to improve the accessibility. Second, a direction set method is used to reduce the complexity of computing clique messages from quartic to cubic. Finally, a coarse-to-fine hierarchical state-space reduction method is presented to decrease redundant states. The results show that a combination of these strategies can greatly accelerate the inference process in large-scale MRFs. For common stereo matching, it results in a speed-up of about 200 times.

[1]  Hans Burkhardt,et al.  Fast Generalized Belief Propagation for MAP Estimation on 2D and 3D Grid-Like Markov Random Fields , 2008, DAGM-Symposium.

[2]  Brian S. R. Armstrong,et al.  Soft Synchronization: Synchronization for Network-Connected Machine Vision Systems , 2007, IEEE Transactions on Industrial Informatics.

[3]  Venkat Chandrasekaran,et al.  Gaussian Multiresolution Models: Exploiting Sparse Markov and Covariance Structure , 2010, IEEE Transactions on Signal Processing.

[4]  Pantelis Georgiadis,et al.  Segmentation of Complementary DNA Microarray Images by Wavelet-Based Markov Random Field Model , 2009, IEEE Transactions on Information Technology in Biomedicine.

[5]  H. Vincent Poor,et al.  How Much Information Can One Get From a Wireless Ad Hoc Sensor Network Over a Correlated Random Field? , 2009, IEEE Transactions on Information Theory.

[6]  William T. Freeman,et al.  Constructing free-energy approximations and generalized belief propagation algorithms , 2005, IEEE Transactions on Information Theory.

[7]  Zhongjie Wang,et al.  Improved Generalized Belief Propagation for Vision Processing , 2011 .

[8]  Dietmar Bruckner,et al.  Behavior Learning in Dwelling Environments With Hidden Markov Models , 2010, IEEE Transactions on Industrial Electronics.

[9]  Christian Wolf,et al.  Document Ink Bleed-Through Removal with Two Hidden Markov Random Fields and a Single Observation Field , 2010, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Pedro F. Felzenszwalb,et al.  Efficient belief propagation for early vision , 2004, CVPR 2004.

[11]  W. Freeman,et al.  Generalized Belief Propagation , 2000, NIPS.

[12]  D. Nistér,et al.  Stereo Matching with Color-Weighted Correlation, Hierarchical Belief Propagation, and Occlusion Handling , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[13]  Jacob Goldberger,et al.  Serial Schedules for Belief-Propagation: Analysis of Convergence Time , 2008, IEEE Transactions on Information Theory.

[14]  Sumana Gupta,et al.  A Fast Nonparametric Noncausal MRF-Based Texture Synthesis Scheme Using a Novel FKDE Algorithm , 2010, IEEE Transactions on Image Processing.

[15]  Mihai Datcu,et al.  Huber–Markov Model for Complex SAR Image Restoration , 2010, IEEE Geoscience and Remote Sensing Letters.

[16]  Ruigang Yang,et al.  Spatial-Temporal Fusion for High Accuracy Depth Maps Using Dynamic MRFs , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[17]  Long Zhu,et al.  Unsupervised Learning of Probabilistic Grammar-Markov Models for Object Categories , 2011, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[18]  Aaas News,et al.  Book Reviews , 1893, Buffalo Medical and Surgical Journal.

[19]  Selin Aviyente,et al.  Prognosis of Gear Failures in DC Starter Motors Using Hidden Markov Models , 2011, IEEE Transactions on Industrial Electronics.

[20]  H. H. Chen,et al.  Video Object Extraction via MRF-Based Contour Tracking , 2010, IEEE Transactions on Circuits and Systems for Video Technology.

[21]  Honghai Liu,et al.  Predicting Future States With $n$ -Dimensional Markov Chains for Fault Diagnosis , 2009, IEEE Transactions on Industrial Electronics.

[22]  Ali Taylan Cemgil,et al.  Gamma Markov Random Fields for Audio Source Modeling , 2009, IEEE Transactions on Audio, Speech, and Language Processing.

[23]  Manjunath V. Joshi,et al.  New Learning Based Super-Resolution: Use of DWT and IGMRF Prior , 2010, IEEE Transactions on Image Processing.