ValueError: 层序列 10 的输入 0 与层不兼容：预期 ndim=5，发现 ndim=4。 Alexnet(cnn) + LSTM

Question

currently making human action recognition to detect a cheating kind on exam from CCTV using AlexNet+LSTM

My Data are raw image in a each class folder with like this

but i got error like this

ValueError: in user code:

/usr/local/lib/python3.7/dist-packages/keras/engine/training.py:830 train_function  *
    return step_function(self, iterator)
/usr/local/lib/python3.7/dist-packages/keras/engine/training.py:813 run_step  *
    outputs = model.train_step(data)
/usr/local/lib/python3.7/dist-packages/keras/engine/training.py:770 train_step  *
    y_pred = self(x, training=True)
/usr/local/lib/python3.7/dist-packages/keras/engine/base_layer.py:989 __call__  *
    input_spec.assert_input_compatibility(self.input_spec, inputs, self.name)
/usr/local/lib/python3.7/dist-packages/keras/engine/input_spec.py:212 assert_input_compatibility  *
    raise ValueError('Input ' + str(input_index) + ' of layer ' +

ValueError: Input 0 of layer sequential_10 is incompatible with the layer: expected ndim=5, found ndim=4. Full shape received: (None, None, None, None)

The error come when i do the model.fit

from what i read it said that the problem at the input_shape but i am still doesnt found the solution of my problem the link of my code at colab can be found here ,

i still dont understand what is the problem, i check the documentation for input_shape in TimeDistributed and it's the same for (timeSteps, height, width, channels)

is it from my ImageDataGenerator?or did i do wrong on making my model?

i would appreciate if anybody have the experience in this matter and try to help my problem

Thank you so much !

Answer 1

As error message says expected dimension 5D but received 4D input. Time distributed layer expects input of shape 5D.

Working sample code:

Consider a batch of 32 video samples, where each sample is a 128x128 RGB image with channels_last data format, across 10 timesteps. The batch input shape is (32, 10, 128, 128, 3) .

inputs = tf.keras.Input(shape=(10, 128, 128, 3))
conv_2d_layer = tf.keras.layers.Conv2D(64, (3, 3))
outputs = tf.keras.layers.TimeDistributed(conv_2d_layer)(inputs)

Output:

TensorShape([None, 10, 126, 126, 64])