层“conv2d”的输入 0 与层不兼容：预期 min_ndim=4，发现 ndim=3。 收到的完整形状：（32、28、28）

Question

I wrote a CNN function for fashion minst dataset, whenever I try to call it in the main, I receive this error: Input 0 of layer "conv2d" is incompatible with the layer: expected min_ndim=4, found ndim=3. Full shape received: (32, 28, 28).

Here's how did I call it:

cnn_model = create_CNN(28, 28, 3, 10)

Here's the CNN code:

def create_CNN(height, weight, channels, classes):
  model = Sequential()
  model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(height, weight, channels)))
  model.add(MaxPooling2D((2, 2)))
  model.add(Conv2D(64, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2)))
  model.add(Flatten())
  model.add(Dense(64, activation='relu'))
  model.add(Dense(classes))

  model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
  
  return model

Answer 1

def create_CNN(classes:int=10, input_shape:tuple=(28, 28, 3)):
  model = Sequential()
  model.add(Conv2D(32, (3, 3), activation='relu', input_shape=input_shape))
  model.add(MaxPooling2D((2, 2)))
  model.add(Conv2D(64, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2)))
  model.add(Flatten())
  model.add(Dense(64, activation='relu'))
  model.add(Dense(classes, activation='softmax'))

  model.compile(optimizer='adam', loss='Sparse_categorical_crossentropy', metrics=['accuracy'])
  
  return model

and also reshape the value, if your data will store in x variable then,

x = data  # your Mnist Fashion Data
print(x.ndim, x.shape)# No. Of Dimensions 3 and Shape is (32, 28, 38)
x = data.reshape((32, 28, 28, 3))  # Reshaping the Data
print(x.ndim, x.shape) # After Reshaping, it No. Of Dimensiona 4 and shape is (32, 28, 28, 3)

I Hope, It will Work Properly.