![]() You need to modify the code as below to make it reusable for printing the Conv3d Convolution kernel in Gray scale. The code you have used is good for displaying the Conv2D Convolution kernel in gray scale OR Conv3D Convolution kernel in Color(because of the depth of kernel causing merge of three gray scale images). ![]() Is this method correct or is there any better way in which this can be done? ![]() # kernel shape 3x3x3 but to plot it converting into 3 3×3 filters # normalize filter values to 0-1 so we can visualize themį_min, f_max = filters.min(), filters.max()įilters = (filters – f_min) / (f_max – f_min) # retrieve weights from the 1st conv layer layerįilters, biases = _weights() Please find below the code I used to plot the weights from keras.models import load_model The conv layer consists of 5 layers #model.add(layers.Conv3D(5, (3, 3, 3), padding=’same’)) Since plotting in 3d is not possible i tried to split the kernels into 3 3×3 for plotting. I am using 3d kernel of size 3x3x3 for the convoution layer and would like to plot the weights of the layer.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |