Not sure how to apply this to 3d Convolutions.

[Dekermanjian]

Hi, first I would like to say I am very new to using your software please forgive me if this is very obvious.

I have a model with two outputs:

I pull the layers and the 1 output I am interested in:

# Convolutional layers indices
ixs = [10]
# Create outputs of each convolutional layer
outputs = [dcec.model.layers[i].output for i in ixs]
# define model
model_grad = Model(inputs=dcec.model.inputs, outputs=outputs)

resulting in:

I create the grad cam object:

# Create GradCAM++ object
gradcam = GradcamPlusPlus(model_grad,
                          model_modifier=ReplaceToLinear(),
                          clone=True)

Then I try to get the cam:

cam = gradcam(CategoricalScore([0,1,2,3,4]),
              tf.expand_dims(x[1], axis=3))

but this results in an InvalidArgumentError:

InvalidArgumentError: slice index 1 of dimension 0 out of bounds. [Op:StridedSlice] name: strided_slice/

I am not sure what the error is trying to indicate. Any help will be greatly appreciated, thank you.