I'm a student and beginner with Keras. I wonder how to use a layer such in the same way of a python function. I want to define a layers, a Dense one, for example, and then be able to compute the image of an x given by the layer.
Thanks and have a good day.
So first, take a look at Keras Functional API. That way you can do something like this:
from keras.layers import *
# load some data into x
x = load()
dense_output = Dense(4096, activation='relu')(x)
dropout_output = Dropout(0.3)(dense_output)
...
You can do basically everything you need with Functional API. I hope this clarifies the situation for you :)