24 #ifndef ARM_COMPUTE_GRAPH_CONVOLUTION_LAYER_NODE_H 25 #define ARM_COMPUTE_GRAPH_CONVOLUTION_LAYER_NODE_H 116 unsigned int _num_groups;
bool forward_descriptors() override
Forwards descriptor information to outputs if possible.
FastMathHint fast_math_hint() const
Fast math hint accessor.
ActivationLayerInfo fused_activation() const
Returns fused activation.
NodeType type() const override
Returns node's type.
Activation Layer Information class.
unsigned int num_groups() const
Number of groups in convolution accessor.
Copyright (c) 2017-2021 Arm Limited.
static constexpr NodeType node_type
Fast math disabled for Convolution layer.
void set_convolution_method(ConvolutionMethod method)
Sets the convolution layer method to use.
Quantization information.
void set_fast_math_hint(FastMathHint hint)
Sets the fast math fast hint.
static TensorDescriptor compute_output_descriptor(const TensorDescriptor &input_descriptor, const TensorDescriptor &weights_descriptor, const PadStrideInfo &info)
Computes convolution output descriptor.
PadStrideInfo convolution_info() const
Convolution metadata accessor.
Padding and stride information class.
FastMathHint
Enable or disable fast math for Convolution layer.
ConvolutionMethod convolution_method() const
Convolution layer method accessor.
ScaleKernelInfo info(interpolation_policy, default_border_mode, PixelValue(), sampling_policy, false)
ConvolutionLayerNode(PadStrideInfo info, unsigned int num_groups=1, ConvolutionMethod method=ConvolutionMethod::Default, FastMathHint fast_math_hint=FastMathHint::Disabled, QuantizationInfo out_quant_info=QuantizationInfo())
Constructor.
ConvolutionMethod
Supported Convolution layer methods.
void set_fused_activation(ActivationLayerInfo fused_activation)
Sets fused activation.
Default approach using internal heuristics.
void accept(INodeVisitor &v) override
Accepts a node visitor.
TensorDescriptor configure_output(size_t idx) const override
Calculates output configuration.