36 std::vector<TensorShape>
InferOutputShapes(
const std::vector<TensorShape> &inputShapes)
const override;
39 void Accept(ILayerVisitor& visitor)
const override;
std::vector< TensorShape > InferOutputShapes(const std::vector< TensorShape > &inputShapes) const override
By default returns inputShapes if the number of inputs are equal to number of outputs, otherwise infers the output shapes from given input shapes and layer properties.
#define ARMNN_NO_DEPRECATE_WARN_BEGIN
virtual std::unique_ptr< IWorkload > CreateWorkload(const IWorkloadFactory &factory) const override
Makes a workload for the Pad type.
~PadLayer()=default
Default destructor.
ARMNN_NO_DEPRECATE_WARN_BEGIN void Accept(ILayerVisitor &visitor) const override
Copyright (c) 2021 ARM Limited and Contributors.
This layer represents a pad operation.
A PadDescriptor for the PadLayer.
#define ARMNN_NO_DEPRECATE_WARN_END
PadLayer(const PadDescriptor ¶m, const char *name)
Constructor to create a PadLayer.
void ValidateTensorShapesFromInputs() override
Check if the input tensor shape(s) will lead to a valid configuration of PadLayer.
PadLayer * Clone(Graph &graph) const override
Creates a dynamically-allocated copy of this layer.