ewen1024
ewen1024
I am not sure about that. But I guess something relevant to spatial_dropout . Did you check that?
@bhatshashank You should stop if you saw the accuracy was 0 after one hour. There must be something wrong. If you only changed the pic folder, make the txt file...
@wzhouuu yeah, directly run the training of encoder-decoder works for me.
dropout is slow which takes about half of the time of the whole iteration. Just replace it with caffe dropout, if you want. @chamcham115
@gjy3035 @peternara Hi I encountered and solved the problem today. First of all, make sure your label is clean enough. LIke I have 2 classes, so all the label pixel...
@sunpeng1996 1.While the weights is calculated from the statistics, you can't blame the distribution is too large when the distribution IS large. 2.Though my code has performance boost after the...
first of all, layer 1_0_3 is the spatial dropout layer implemented by cpu. plz make sure you are doing right with the python script
The author says in the segnet repo that 6.0 is not faster, so he has no plan for that.