I am currently writing the code, the model to take 4 images as an input. All 4 images are passed through the convolution network and then all image's output is concatenated. But while compiling I am getting an error of "Input 0 of layer max_pooling2d_111 is incompatible with the layer: expected ndim=4 found ndim=5. Full shape received: [None, 4, 128, 128, 16]"
I know there is 1 more dimension in the input (size of 4) but how should I do this or solve this issue?
My model code is :
#Build the model
inputs1 = tf.keras.layers.Input((IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS))
o1 = tf.keras.layers.Lambda(lambda x: x / 255)(inputs1)
c1 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(o1)
c1 = tf.keras.layers.Dropout(0.1)(c1)
c1 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c1)
p1 = tf.keras.layers.MaxPooling2D((2, 2))(c1)
#p1 = c1
c1 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p1)
c1 = tf.keras.layers.Dropout(0.1)(c1)
c1 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c1)
p1 = tf.keras.layers.MaxPooling2D((2, 2))(c1)
#p1 = c1
c1 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p1)
c1 = tf.keras.layers.Dropout(0.2)(c1)
c1 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c1)
p1 = tf.keras.layers.MaxPooling2D((2, 2))(c1)
#p1 = c1
c1 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p1)
c1 = tf.keras.layers.Dropout(0.2)(c1)
c1 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c1)
p1 = tf.keras.layers.MaxPooling2D(pool_size=(2, 2))(c1)
#p1 = c1
inputs2 = tf.keras.layers.Input((IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS))
o2 = tf.keras.layers.Lambda(lambda x: x / 255)(inputs2)
c2 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(o2)
c2 = tf.keras.layers.Dropout(0.1)(c2)
c2 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c2)
p2 = tf.keras.layers.MaxPooling2D((2, 2))(c2)
#p2 = c2
c2 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p2)
c2 = tf.keras.layers.Dropout(0.1)(c2)
c2 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c2)
p2 = tf.keras.layers.MaxPooling2D((2, 2))(c2)
#p2 = c2
c2 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p2)
c2 = tf.keras.layers.Dropout(0.2)(c2)
c2 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c2)
p2 = tf.keras.layers.MaxPooling2D((2, 2))(c2)
#p2 = c2
c2 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p2)
c2 = tf.keras.layers.Dropout(0.2)(c2)
c2 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c2)
p2 = tf.keras.layers.MaxPooling2D(pool_size=(2, 2))(c2)
#p2 = c2
inputs3 = tf.keras.layers.Input((IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS))
o3 = tf.keras.layers.Lambda(lambda x: x / 255)(inputs3)
c3 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(o3)
c3 = tf.keras.layers.Dropout(0.1)(c3)
c3 = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c3)
p3 = tf.keras.layers.MaxPooling2D((2, 2))(c3)
#p3 = c3
c3 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p3)
c3 = tf.keras.layers.Dropout(0.1)(c3)
c3 = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c3)
p3 = tf.keras.layers.MaxPooling2D((2, 2))(c3)
#p3 = c3
c3 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p3)
c3 = tf.keras.layers.Dropout(0.2)(c3)
c3 = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c3)
p3 = tf.keras.layers.MaxPooling2D((2, 2))(c3)
#p3 = c3
c3 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p3)
c3 = tf.keras.layers.Dropout(0.2)(c3)
c3 = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c3)
p3 = tf.keras.layers.MaxPooling2D(pool_size=(2, 2))(c3)
#p3 = c3
P = tf.keras.layers.concatenate([p1, p2, p3])
cp = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(P)
cp = tf.keras.layers.Dropout(0.3)(cp)
cp = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(cp)
#Build the model
inputs = tf.keras.layers.Input((IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS))
s = tf.keras.layers.Lambda(lambda x: x / 255)(inputs)
c = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(s)
c = tf.keras.layers.Dropout(0.1)(c)
c = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c)
p = tf.keras.layers.MaxPooling2D((2, 2))(c)
#p = c
c = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p)
c = tf.keras.layers.Dropout(0.1)(c)
c = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c)
p = tf.keras.layers.MaxPooling2D((2, 2))(c)
#p = c
c = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p)
c = tf.keras.layers.Dropout(0.2)(c)
c = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c)
p = tf.keras.layers.MaxPooling2D((2, 2))(c)
#p = c
c = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p)
c = tf.keras.layers.Dropout(0.2)(c)
c = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c)
p = tf.keras.layers.MaxPooling2D(pool_size=(2, 2))(c)
#p = c
c = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(p)
c = tf.keras.layers.Dropout(0.3)(c)
c = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(c)
one = tf.keras.layers.concatenate([cp,c])
#one = tf.keras.layers.add([cp,c])
#Compression
onec = tf.keras.layers.Conv2D(512, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(one)
onec = tf.keras.layers.Dropout(0.3)(onec)
onec = tf.keras.layers.Conv2D(512, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(onec)
onep = tf.keras.layers.MaxPooling2D(pool_size=(2, 2))(onec)
#Decompression
d = tf.keras.layers.Conv2DTranspose(256, (2, 2), strides=(2, 2), padding='same')(onep)
d = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Dropout(0.3)(d)
d = tf.keras.layers.Conv2D(256, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Conv2DTranspose(128, (2, 2), strides=(2, 2), padding='same')(d)
d = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Dropout(0.3)(d)
d = tf.keras.layers.Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Conv2DTranspose(64, (2, 2), strides=(2, 2), padding='same')(d)
d = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Dropout(0.2)(d)
d = tf.keras.layers.Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Conv2DTranspose(332, (2, 2), strides=(2, 2), padding='same')(d)
d = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Dropout(0.1)(d)
d = tf.keras.layers.Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Conv2DTranspose(16, (2, 2), strides=(2, 2), padding='same')(d)
d = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
d = tf.keras.layers.Dropout(0.1)(d)
d = tf.keras.layers.Conv2D(16, (3, 3), activation='relu', kernel_initializer='he_normal', padding='same')(d)
outputs = tf.keras.layers.Conv2D(1, (1, 1), activation='relu',name = "ouput")(d)
model = tf.keras.Model(inputs=[inputs,inputs1,inputs2,inputs3], outputs=[outputs])
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
Anyone can please help me or suggest to me how to do it.