Input Error

[elJonathan]

I am getting error in the code for inputs. Please let me know the dimensions of the images that you are using as I am getting the following error. ValueError: Error when checking input: expected input to have 5 dimensions, but got array with shape(160, 1)

[RishalAggarwal]

It seems that you may not be providing a 3D images as input. Typically the network requires a 5 dimensional input: 3 dimensions of the image, 1 dimension for channel and 1 dimension for the batch size. I don't remember exactly what the size of each dimension was, but this code is adaptable to different sizes as long as the number of dimensions are valid.

[MohammadSalehi72]

Hello Rishal,

I have a problem with the input shape for the model. According to the "training cnn.py" script, my input images are :(12, 106, 106, 120) in shape. Next, in "tflearn.input_data(shape=[None, sh[1], sh[2], sh[3], sh[4]])" layer, I don't understand 4th dimension. And is my image's shape correct?

can you please explain that?

[RishalAggarwal]

You need to add a dimension to the images which will correspond to the channels (analogous to RGB in a regular 2D image)

[dilipv09]

Hi Rishal,

My input shape is ((1, 96, 96, 60),3). model=ResnetBuilder.build_resnet_34((1,96,96,60),3). Facing below error.

Traceback (most recent call last): File "train_resnet.py", line 269, in model=ResnetBuilder.build_resnet_34((1,96,96,60),3) File "train_resnet.py", line 255, in build_resnet_34 return ResnetBuilder.build(input_shape, num_outputs, basic_block, [3, 4, 6, 3]) File "train_resnet.py", line 233, in build block = _residual_block(block_fn, filters=filters, repetitions=r, is_first_layer=(i == 0))(block) File "train_resnet.py", line 120, in f is_first_block_of_first_layer=(is_first_layer and i == 0))(input) File "train_resnet.py", line 144, in f return _shortcut(input, residual) File "train_resnet.py", line 106, in _shortcut kernel_regularizer=l2(0.0001))(input) File "/data/users/dverma01/conda/envs/gcolab/lib/python3.6/site-packages/tensorflow/python/keras/engine/base_layer.py", line 897, in call self._maybe_build(inputs) File "/data/users/dverma01/conda/envs/gcolab/lib/python3.6/site-packages/tensorflow/python/keras/engine/base_layer.py", line 2416, in _maybe_build self.build(input_shapes) # pylint:disable=not-callable File "/data/users/dverma01/conda/envs/gcolab/lib/python3.6/site-packages/tensorflow/python/keras/layers/convolutional.py", line 190, in build data_format=self._conv_op_data_format) File "/data/users/dverma01/conda/envs/gcolab/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py", line 1066, in init num_spatial_dims, strides, dilation_rate) File "/data/users/dverma01/conda/envs/gcolab/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py", line 755, in _get_strides_and_dilation_rate raise ValueError("all values of strides must be positive") ValueError: all values of strides must be positive

Any input is highly appreciated. Thanks!