lucasb-eyer / pydensecrf

Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.
MIT License
1.96k stars 417 forks source link

apply densecrf to node classification #106

Closed tanjia123456 closed 3 years ago

tanjia123456 commented 3 years ago

Hello, thank you for your work I use graph neural network to parcellation brain, now I want to use densecrf as a postprocess. I have get the logits(10242,36) the 10242 is node number, 36 is class number. I also have standard label(10242). So, could you please give me some suggestion about how to add densecrf as postprocess?