Use of the MaxUnPooling2D

General Discussion
,
#1

i want use MaxUnPooling2D in an autoencoder like this code
input_img = Input(shape=(nr_col, nr_rows, 1))
ec_conv1 = Conv2D(N[0], (3,3), activation=‘relu’, padding=‘same’)(input_img)
ec_ouput = MaxPooling2D((2, 2), padding=‘same’)(ec_conv1)
z=Flatten(name=‘VC’)(ec_ouput)
r=Reshape((24,24,N[0]),name=‘Resh’)(z)
dec_conv1 = Conv2D(N[0], (3, 3), activation=‘relu’, padding=‘same’)(r)
dec_upsample1 = tfa.layers.MaxUnpooling2D((2, 2), padding=‘SAME’)(dec_conv1)
dec_output = Conv2D(1, (3,3), activation=‘sigmoid’, padding=‘same’)(dec_upsample1)
autoencoder = Model(input_img, dec_output)
autoencoder.summary()
autoencoder.compile(optimizer=‘adam’, loss=‘binary_crossentropy’, metrics = [‘acc’]) #binary_crossentropy
autoencoder_history = autoencoder.fit(x_train, x_train,batch_size=batch_size,epochs=epochs, verbose=1,validation_data=(x_val, x_val))

But i get an this error
TypeError: in user code: TypeError: tf__call() missing 1 required positional argument: 'mask'

Any help ?

#3

Please update TF Addons version to the last One

#5

did you figure this out? I’m having the same problem