TY - JOUR
T1 - Bayesian image modeling by means of a generalized sparse prior and loopy belief propagation
AU - Tanaka, Kazuyuki
AU - Yasuda, Muneki
AU - Titterington, D. Michael
PY - 2012/11
Y1 - 2012/11
N2 - Bayesian image modeling is presented based on a generalized sparse prior probability distribution. Our prior includes sparsity in each interaction term between every pair of neighbouring pixels in Markov random fields. A new scheme for hyperparameter estimation is based on the conditional maximization of entropy in our generalized sparse prior. In addition, the criterion used for defining the optimal value for sparseness in interactions is that of the maximization of marginal likelihood. Our practical algorithm is based on loopy belief propagation.
AB - Bayesian image modeling is presented based on a generalized sparse prior probability distribution. Our prior includes sparsity in each interaction term between every pair of neighbouring pixels in Markov random fields. A new scheme for hyperparameter estimation is based on the conditional maximization of entropy in our generalized sparse prior. In addition, the criterion used for defining the optimal value for sparseness in interactions is that of the maximization of marginal likelihood. Our practical algorithm is based on loopy belief propagation.
KW - Bayesian statistics
KW - Belief propagation
KW - Markov random fields
KW - Maximum likelihood estimation
KW - Probabilistic image processing
KW - Statistical-mechanical informatics
UR - http://www.scopus.com/inward/record.url?scp=84870184026&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84870184026&partnerID=8YFLogxK
U2 - 10.1143/JPSJ.81.114802
DO - 10.1143/JPSJ.81.114802
M3 - Article
AN - SCOPUS:84870184026
SN - 0031-9015
VL - 81
JO - Journal of the Physical Society of Japan
JF - Journal of the Physical Society of Japan
IS - 11
M1 - 114802
ER -