How to Make a Convolution Kernel in a Layer Not Participate in Back Propagation in Convolutional Networks？

1

I’m doing research on model pruning.
Filter pruning.(Kernel pruning)
I set the value of some filters to zero
then retrain model. But these values become non-zero in the process of Back Propagation.

%E5%BE%AE%E4%BF%A1%E6%88%AA%E5%9B%BE_20190329154417
%E5%BE%AE%E4%BF%A1%E6%88%AA%E5%9B%BE_201903291544171058×269 48.6 KB

What can I do to solve this problem? （Keep these filters at 0 during retraining）

Thanks for your help~~