ValueError: The shape of the input to "Flatten" is not fully defined . Make sure to pass a complete "input_shape" or "batch_input

[AgrawalAmey]

I am trying to run the classification/mvcnn_FPS.py script with Keras 1.2.2 (the one specified in the requirements.txt) but the execution fails with the following error.

Using Theano backend.
Using cuDNN version 5110 on context None
Mapped name None to device cuda: GeForce GTX 1070 (0000:01:00.0)
Traceback (most recent call last):
  File "mvcnn_FPS.py", line 224, in <module>
    vggmodel = load_model_vgg()
  File "mvcnn_FPS.py", line 33, in load_model_vgg
    x = Flatten(name='flatten')(x)
  File "/usr/local/lib/python2.7/dist-packages/keras/engine/topology.py", line 572, in __call__
    self.add_inbound_node(inbound_layers, node_indices, tensor_indices)
  File "/usr/local/lib/python2.7/dist-packages/keras/engine/topology.py", line 635, in add_inbound_node
    Node.create_node(self, inbound_layers, node_indices, tensor_indices)
  File "/usr/local/lib/python2.7/dist-packages/keras/engine/topology.py", line 170, in create_node
    output_shapes = to_list(outbound_layer.get_output_shape_for(input_shapes[0]))
  File "/usr/local/lib/python2.7/dist-packages/keras/layers/core.py", line 476, in get_output_shape_for
    '(got ' + str(input_shape[1:]) + '. '
ValueError: The shape of the input to "Flatten" is not fully defined (got (512, 7, 0). Make sure to pass a complete "input_shape" or "batch_input
_shape" argument to the first layer in your model.

[bradyz]

I didn't write this one so I'm not quite sure - have you tried models in the model directory?

[AgrawalAmey]

Okay, I would try those. Thanks.