Post by Yu Liu » Tue Oct 03, 2017, 19:56
The processing of image seems interesting to me, too, for I've studied image processing for several months in my bachelor time. However for now, I'm learning and implementing CRF, kind of probabilistic graphical model (pgm), to complete my doctoral dissertation on user trust prediction in online social networks. So, I'm focusing on extending DGM to support more general graph building, such as graph with different kinds of nodes and edges, as well as training and testing a batch of images (or other graphs) with features by either SGD or L-BFGS.
My plan is to make my extending compatible with your codes after I've finished my research that, I guess, will be done in one or two months. I hope we can write a paper on DGM, a demo of pgm, together and submit to some conference, if my extending could be approved and you're agree on our collaboration.
I'll read the two papers you mentioned some time later. And questions, if any, will be posted in forum.
The processing of image seems interesting to me, too, for I've studied image processing for several months in my bachelor time. However for now, I'm learning and implementing CRF, kind of probabilistic graphical model (pgm), to complete my doctoral dissertation on user trust prediction in online social networks. So, I'm focusing on extending DGM to support more general graph building, such as graph with different kinds of nodes and edges, as well as training and testing a batch of images (or other graphs) with features by either SGD or L-BFGS.
My plan is to make my extending compatible with your codes after I've finished my research that, I guess, will be done in one or two months. I hope we can write a paper on DGM, a demo of pgm, together and submit to some conference, if my extending could be approved and you're agree on our collaboration.
I'll read the two papers you mentioned some time later. And questions, if any, will be posted in forum.