search

lucasb-eyer / pydensecrf

Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.
MIT License
1.94k stars 414 forks source link

Help with wrapping learning #24

Open ajdroid opened 7 years ago

ajdroid commented 7 years ago

Hey, I'd like to use the learning part of the code and might as well come up with a wrapper. Any pointers on how/where to start?

lucasb-eyer commented 7 years ago

From the wrapper side, I believe it should be enough to wrap more methods in densecrf.pxd, specifically this one.

The reason I never did this myself is I don't know how to actually use this for doing learning and how to test it. So you'll probably spend more time on writing an example for this.

I'd greatly appreciate a PR including an example!

manishh commented 5 years ago

@ajdroid - Did you manage to wrap the learning part of this code? Do you have link to any such repositories?

I want to explore the learning approach for a certain image denosing problem, so any pointers to learning here would be appreciated.