Is there an error in this statement about blob dimensions in the tutorial documentation?

[GoodDeeds]

In the tutorial documentation about Blobs, Layers, and Nets: anatomy of a Caffe model, there is a paragraph:

Parameter blob dimensions vary according to the type and configuration of the layer. For a convolution layer with 96 filters of 11 x 11 spatial dimension and 3 inputs the blob is 96 x 3 x 11 x 11. For an inner product / fully-connected layer with 1000 output channels and 1024 input channels the parameter blob is 1000 x 1024

Shouldn't this be 3 x 96 x 11 x 11 and 1024 x 1000 respectively? Doesn't the batch size come first, followed by the number of filters, and then the image dimensions, in a convolutional layer?

[Przemek D]

Except there is no notion of "batch size" for layer parameters like it is for images, since layers only have one parameter (plus biases, optionally), it is exactly as you think it is: number of filters (96) and then image dimensions (3x11x11) => 96x3x11x11.
Same with InnerProduct: number of "filters" (outputs), 1000, then input width (1024) => 1000x1024.

[GoodDeeds]

Thank you, I was getting confused between data blobs and parameter blobs.