AggregationLayer

represents a layer that aggregates a tensor of arbitrary rank into a vector, using the function f.

aggregates a tensor at level n.

aggregates a tensor at levels n₁ through n₂.

AggregationLayer[f,{n₁,n₂,…}]

aggregates a tensor at levels n₁,n₂,….

Details and Options

AggregationLayer[f] operates on a tensor of dimensions {d₁,…,d_n} to produce a vector of size d₁, effectively mapping the function f over a list of flattened subtensors of size d₂×…×d_n.
AggregationLayer[f,{n₁,…,n_k}] operates on a rank-m tensor of dimensions {d₁,…,d_m} to produce a tensor of rank m-k, with dimensions Complement[{d₁,…,d_m},{d_n₁,…,d_{n_k}}].
Possible values for f are Mean, Min, Max, Total and Times.
AggregationLayer[f] is equivalent to AggregationLayer[f,2;;All].
AggregationLayer[…][input] explicitly computes the output from applying the layer.
AggregationLayer[…][{input₁,input₂,…}] explicitly computes outputs for each of the input_i.
AggregationLayer is typically used inside NetChain, NetGraph, etc. as the final stage in a chain of convolutions, poolings, etc. to convert a tensor with spatial dimensions into a fixed-size vector representation.
AggregationLayer can operate on tensors that contain "Varying" dimensions.
AggregationLayer exposes the following ports for use in NetGraph etc.:
"Input" a tensor

"Output" a tensor
When it cannot be inferred from other layers in a larger net, the option "Input"{d₁,…,d_n} can be used to fix the input of AggregationLayer to be a tensor of dimensions {d₁,…,d_n}.