Re: [Dclib-devel] dnn_metric_learning_on_images_ex - can't load trained model
Brought to you by:
davisking
From: Ben G. <ben...@gm...> - 2018-05-23 16:36:29
|
OK, I added my code to examples CMakeFile and its working (need to understand why when I build it manually it doesn't works??) Anyways, How can I get the metric from my network to see the prediction? I trained my model for two classes and I'm executes it like this: std::vector<matrix<float,0,1>> embedded = net(faces); how can I fetch the results metric for each face from 'embedded'? Regards, Ben 2018-05-22 22:19 GMT+03:00 Ben Goz <ben...@gm...>: > I'm using google test framework with empty Setup function and this is the > only code I'm using on my test item: > anet_type t; > cout << "Trying loading net" << endl; > deserialize("metric_network_renset.dat") >> t; > cout << "Loaded net" << endl; > > nothing else before and after and I'm getting memory corruption all the > time. > Can't understand what could go wrong?!?! > > 2018-05-22 22:00 GMT+03:00 Davis King <dav...@gm...>: > >> No, you don't need to do anything. The things you are describing should >> be impossible, which means you have probably clobbered your function stack >> or done something similarly bad in other parts of your code. You just need >> to debug your program. Look for something bad that happens prior to this. >> This is just the symptom of some bug somewhere else, not the cause. >> >> You can load it with anet_type or net_type. Both are fine. >> >> ------------------------------------------------------------ >> ------------------ >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> Dclib-devel mailing list >> Dcl...@li... >> https://lists.sourceforge.net/lists/listinfo/dclib-devel >> >> > |