How to write a custom function in keras without using backend functions?

Question

I want to get the output of a keras layer which is a tensor of images, convert it to numpy arrays and apply a custom function on them without using backend functions, and return the output to the model. So it would be something like this:

def sum(args):
    # convert args to numpy arrays
    ???
    #calculating sum of all the elements
    s=np.sum(args)
    return s

Thank you very much for your help.

MTMD · Accepted Answer · 2018-08-31 00:42:26Z

1

You can use the following format using Keras on Tensorflow:

  x = layers.Conv2D(32, (3, 3), strides=(1, 1), 
    padding='same', name='conv_1', use_bias=False)(input_image)
  x = my_function(x)
  x = layers.BatchNormalization(name='norm_1')(x)

answered Aug 31, 2018 at 0:42

MTMD

1,2522 gold badges13 silver badges32 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Iman Sajedian Over a year ago

It doesn't work like that. I already tried it. InvalidArgumentError: You must feed a value for placeholder tensor 'input_19' with dtype float and shape [?,100] [[Node: input_19 = Placeholder[dtype=DT_FLOAT, shape=[?,100], _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

Collectives™ on Stack Overflow

How to write a custom function in keras without using backend functions?

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related