diff options
Diffstat (limited to 'src')
58 files changed, 912 insertions, 618 deletions
diff --git a/src/armnn/InternalTypes.cpp b/src/armnn/InternalTypes.cpp index 7c39128bec..f713644656 100644 --- a/src/armnn/InternalTypes.cpp +++ b/src/armnn/InternalTypes.cpp @@ -20,6 +20,7 @@ char const* GetLayerTypeAsCString(LayerType type) case LayerType::ArgMinMax: return "ArgMinMax"; case LayerType::BatchNormalization: return "BatchNormalization"; case LayerType::BatchToSpaceNd: return "BatchToSpaceNd"; + case LayerType::Comparison: return "Comparison"; case LayerType::Concat: return "Concat"; case LayerType::Constant: return "Constant"; case LayerType::ConvertFp16ToFp32: return "ConvertFp16ToFp32"; @@ -31,12 +32,10 @@ char const* GetLayerTypeAsCString(LayerType type) case LayerType::Dequantize: return "Dequantize"; case LayerType::DetectionPostProcess: return "DetectionPostProcess"; case LayerType::Division: return "Division"; - case LayerType::Equal: return "Equal"; case LayerType::FakeQuantization: return "FakeQuantization"; case LayerType::Floor: return "Floor"; case LayerType::FullyConnected: return "FullyConnected"; case LayerType::Gather: return "Gather"; - case LayerType::Greater: return "Greater"; case LayerType::Input: return "Input"; case LayerType::InstanceNormalization: return "InstanceNormalization"; case LayerType::L2Normalization: return "L2Normalization"; diff --git a/src/armnn/InternalTypes.hpp b/src/armnn/InternalTypes.hpp index 895fe3235d..d7f932f699 100644 --- a/src/armnn/InternalTypes.hpp +++ b/src/armnn/InternalTypes.hpp @@ -20,6 +20,7 @@ enum class LayerType ArgMinMax, BatchNormalization, BatchToSpaceNd, + Comparison, Concat, Constant, ConvertFp16ToFp32, @@ -31,12 +32,10 @@ enum class LayerType Dequantize, DetectionPostProcess, Division, - Equal, FakeQuantization, Floor, FullyConnected, Gather, - Greater, Input, InstanceNormalization, L2Normalization, diff --git a/src/armnn/LayerSupport.cpp b/src/armnn/LayerSupport.cpp index f88e4e1cc9..997b5f245a 100644 --- a/src/armnn/LayerSupport.cpp +++ b/src/armnn/LayerSupport.cpp @@ -262,7 +262,12 @@ bool IsEqualSupported(const BackendId& backend, char* reasonIfUnsupported, size_t reasonIfUnsupportedMaxLength) { - FORWARD_LAYER_SUPPORT_FUNC(backend, IsEqualSupported, input0, input1, output); + FORWARD_LAYER_SUPPORT_FUNC(backend, + IsComparisonSupported, + input0, + input1, + output, + ComparisonDescriptor(ComparisonOperation::Equal)); } bool IsFakeQuantizationSupported(const BackendId& backend, @@ -317,7 +322,12 @@ bool IsGreaterSupported(const BackendId& backend, char* reasonIfUnsupported, size_t reasonIfUnsupportedMaxLength) { - FORWARD_LAYER_SUPPORT_FUNC(backend, IsGreaterSupported, input0, input1, output); + FORWARD_LAYER_SUPPORT_FUNC(backend, + IsComparisonSupported, + input0, + input1, + output, + ComparisonDescriptor(ComparisonOperation::Greater)); } bool IsInputSupported(const BackendId& backend, diff --git a/src/armnn/LayersFwd.hpp b/src/armnn/LayersFwd.hpp index 7bb9c64818..6c30749904 100644 --- a/src/armnn/LayersFwd.hpp +++ b/src/armnn/LayersFwd.hpp @@ -12,6 +12,7 @@ #include "layers/ArgMinMaxLayer.hpp" #include "layers/BatchNormalizationLayer.hpp" #include "layers/BatchToSpaceNdLayer.hpp" +#include "layers/ComparisonLayer.hpp" #include "layers/ConcatLayer.hpp" #include "layers/ConstantLayer.hpp" #include "layers/ConvertFp16ToFp32Layer.hpp" @@ -23,12 +24,10 @@ #include "layers/DequantizeLayer.hpp" #include "layers/DetectionPostProcessLayer.hpp" #include "layers/DivisionLayer.hpp" -#include "layers/EqualLayer.hpp" #include "layers/FakeQuantizationLayer.hpp" #include "layers/FloorLayer.hpp" #include "layers/FullyConnectedLayer.hpp" #include "layers/GatherLayer.hpp" -#include "layers/GreaterLayer.hpp" #include "layers/InputLayer.hpp" #include "layers/InstanceNormalizationLayer.hpp" #include "layers/L2NormalizationLayer.hpp" @@ -97,6 +96,7 @@ DECLARE_LAYER(Addition) DECLARE_LAYER(ArgMinMax) DECLARE_LAYER(BatchNormalization) DECLARE_LAYER(BatchToSpaceNd) +DECLARE_LAYER(Comparison) DECLARE_LAYER(Concat) DECLARE_LAYER(Constant) DECLARE_LAYER(ConvertFp16ToFp32) @@ -108,12 +108,10 @@ DECLARE_LAYER(DepthwiseConvolution2d) DECLARE_LAYER(Dequantize) DECLARE_LAYER(DetectionPostProcess) DECLARE_LAYER(Division) -DECLARE_LAYER(Equal) DECLARE_LAYER(FakeQuantization) DECLARE_LAYER(Floor) DECLARE_LAYER(FullyConnected) DECLARE_LAYER(Gather) -DECLARE_LAYER(Greater) DECLARE_LAYER(Input) DECLARE_LAYER(InstanceNormalization) DECLARE_LAYER(L2Normalization) diff --git a/src/armnn/Network.cpp b/src/armnn/Network.cpp index b2fc1a6389..857f6b3959 100644 --- a/src/armnn/Network.cpp +++ b/src/armnn/Network.cpp @@ -938,6 +938,12 @@ IConnectableLayer* Network::AddBatchToSpaceNdLayer(const BatchToSpaceNdDescripto return m_Graph->AddLayer<BatchToSpaceNdLayer>(batchToSpaceNdDescriptor, name); } +IConnectableLayer* Network::AddComparisonLayer(const ComparisonDescriptor& comparisonDescriptor, + const char* name) +{ + return m_Graph->AddLayer<ComparisonLayer>(comparisonDescriptor, name); +} + IConnectableLayer* Network::AddFullyConnectedLayerImpl(const FullyConnectedDescriptor& fullyConnectedDescriptor, const ConstTensor& weights, const Optional<ConstTensor>& biases, @@ -1436,12 +1442,12 @@ IConnectableLayer* Network::AddStridedSliceLayer(const StridedSliceDescriptor& s IConnectableLayer* Network::AddGreaterLayer(const char* name) { - return m_Graph->AddLayer<GreaterLayer>(name); + return AddComparisonLayer(ComparisonDescriptor(ComparisonOperation::Greater), name); } IConnectableLayer* Network::AddEqualLayer(const char* name) { - return m_Graph->AddLayer<EqualLayer>(name); + return AddComparisonLayer(ComparisonDescriptor(ComparisonOperation::Equal), name); } IConnectableLayer* Network::AddRsqrtLayer(const char * name) diff --git a/src/armnn/Network.hpp b/src/armnn/Network.hpp index ad1e7c456e..c1d99a9010 100644 --- a/src/armnn/Network.hpp +++ b/src/armnn/Network.hpp @@ -42,6 +42,9 @@ public: IConnectableLayer* AddBatchToSpaceNdLayer(const BatchToSpaceNdDescriptor& batchToSpaceNdDescriptor, const char* name = nullptr) override; + IConnectableLayer* AddComparisonLayer(const ComparisonDescriptor& comparisonDescriptor, + const char* name = nullptr) override; + IConnectableLayer* AddConcatLayer(const ConcatDescriptor& concatDescriptor, const char* name = nullptr) override; @@ -197,8 +200,10 @@ public: IConnectableLayer* AddMinimumLayer(const char* name = nullptr) override; + ARMNN_DEPRECATED_MSG("Use AddComparisonLayer instead") IConnectableLayer* AddGreaterLayer(const char* name = nullptr) override; + ARMNN_DEPRECATED_MSG("Use AddComparisonLayer instead") IConnectableLayer* AddEqualLayer(const char* name = nullptr) override; IConnectableLayer* AddRsqrtLayer(const char* name = nullptr) override; diff --git a/src/armnn/layers/ComparisonLayer.cpp b/src/armnn/layers/ComparisonLayer.cpp new file mode 100644 index 0000000000..75518e580e --- /dev/null +++ b/src/armnn/layers/ComparisonLayer.cpp @@ -0,0 +1,80 @@ +// +// Copyright © 2019 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// + +#include "ComparisonLayer.hpp" + +#include "LayerCloneBase.hpp" + +#include <backendsCommon/WorkloadData.hpp> +#include <backendsCommon/WorkloadFactory.hpp> + +#include <algorithm> + +namespace armnn +{ + +ComparisonLayer::ComparisonLayer(const ComparisonDescriptor& param, const char* name) + : LayerWithParameters(2, 1, LayerType::Comparison, param, name) +{ +} + +std::unique_ptr<IWorkload> ComparisonLayer::CreateWorkload(const Graph& graph, + const IWorkloadFactory& factory) const +{ + ComparisonQueueDescriptor descriptor; + return factory.CreateComparison(descriptor, PrepInfoAndDesc(descriptor, graph)); +} + +ComparisonLayer* ComparisonLayer::Clone(Graph& graph) const +{ + return CloneBase<ComparisonLayer>(graph, m_Param, GetName()); +} + +std::vector<TensorShape> ComparisonLayer::InferOutputShapes(const std::vector<TensorShape>& inputShapes) const +{ + BOOST_ASSERT(inputShapes.size() == 2); + const TensorShape& input0 = inputShapes[0]; + const TensorShape& input1 = inputShapes[1]; + + BOOST_ASSERT(input0.GetNumDimensions() == input1.GetNumDimensions()); + unsigned int numDims = input0.GetNumDimensions(); + + std::vector<unsigned int> dims(numDims); + for (unsigned int i = 0; i < numDims; i++) + { + unsigned int dim0 = input0[i]; + unsigned int dim1 = input1[i]; + + BOOST_ASSERT_MSG(dim0 == dim1 || dim0 == 1 || dim1 == 1, + "Dimensions should either match or one should be of size 1."); + + dims[i] = std::max(dim0, dim1); + } + + return std::vector<TensorShape>({ TensorShape(numDims, dims.data()) }); +} + +void ComparisonLayer::ValidateTensorShapesFromInputs() +{ + VerifyLayerConnections(2, CHECK_LOCATION()); + + std::vector<TensorShape> inferredShapes = InferOutputShapes({ + GetInputSlot(0).GetConnection()->GetTensorInfo().GetShape(), + GetInputSlot(1).GetConnection()->GetTensorInfo().GetShape() + }); + BOOST_ASSERT(inferredShapes.size() == 1); + + ConditionalThrowIfNotEqual<LayerValidationException>( + "ComparisonLayer: TensorShape set on OutputSlot[0] does not match the inferred shape.", + GetOutputSlot(0).GetTensorInfo().GetShape(), + inferredShapes[0]); +} + +void ComparisonLayer::Accept(ILayerVisitor& visitor) const +{ + visitor.VisitComparisonLayer(this, GetParameters(), GetName()); +} + +} // namespace armnn diff --git a/src/armnn/layers/ComparisonLayer.hpp b/src/armnn/layers/ComparisonLayer.hpp new file mode 100644 index 0000000000..bbc2b573bf --- /dev/null +++ b/src/armnn/layers/ComparisonLayer.hpp @@ -0,0 +1,50 @@ +// +// Copyright © 2019 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// + +#pragma once + +#include "LayerWithParameters.hpp" + +namespace armnn +{ + +/// This layer represents a comparison operation. +class ComparisonLayer : public LayerWithParameters<ComparisonDescriptor> +{ +public: + /// Makes a workload for the Comparison type + /// @param [in] graph The graph where this layer can be found + /// @param [in] factory The workload factory which will create the workload + /// @return A pointer to the created workload, or nullptr if not created + virtual std::unique_ptr<IWorkload> CreateWorkload(const Graph& graph, + const IWorkloadFactory& factory) const override; + + /// Creates a dynamically-allocated copy of this layer + /// @param [in] graph The graph into which this layer is being cloned + ComparisonLayer* Clone(Graph& graph) 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. + /// @param [in] inputShapes The input shapes layer has. + /// @return A vector to the inferred output shape. + std::vector<TensorShape> InferOutputShapes(const std::vector<TensorShape>& inputShapes) const override; + + /// Check if the input tensor shape(s) will lead to a valid configuration + /// of @ref ComparisonLayer + void ValidateTensorShapesFromInputs() override; + + void Accept(ILayerVisitor& visitor) const override; + +protected: + /// Constructor to create a ComparisonLayer + /// @param [in] param ComparisonDescriptor to configure the ComparisonLayer + /// @param [in] name Optional name for the layer + ComparisonLayer(const ComparisonDescriptor& param, const char* name); + + /// Default destructor + ~ComparisonLayer() = default; +}; + +} // namespace armnn diff --git a/src/armnn/layers/EqualLayer.cpp b/src/armnn/layers/EqualLayer.cpp deleted file mode 100644 index 7d16668b06..0000000000 --- a/src/armnn/layers/EqualLayer.cpp +++ /dev/null @@ -1,39 +0,0 @@ -// -// Copyright © 2017 Arm Ltd. All rights reserved. -// SPDX-License-Identifier: MIT -// - -#include "EqualLayer.hpp" - -#include "LayerCloneBase.hpp" - -#include <armnn/TypesUtils.hpp> -#include <backendsCommon/WorkloadData.hpp> -#include <backendsCommon/WorkloadFactory.hpp> - -namespace armnn -{ - -EqualLayer::EqualLayer(const char* name) - : ElementwiseBaseLayer(2, 1, LayerType::Equal, name) -{ -} - -std::unique_ptr<IWorkload> EqualLayer::CreateWorkload(const Graph& graph, - const IWorkloadFactory& factory) const -{ - EqualQueueDescriptor descriptor; - return factory.CreateEqual(descriptor, PrepInfoAndDesc(descriptor, graph)); -} - -EqualLayer* EqualLayer::Clone(Graph& graph) const -{ - return CloneBase<EqualLayer>(graph, GetName()); -} - -void EqualLayer::Accept(ILayerVisitor& visitor) const -{ - visitor.VisitEqualLayer(this, GetName()); -} - -} // namespace armnn diff --git a/src/armnn/layers/EqualLayer.hpp b/src/armnn/layers/EqualLayer.hpp deleted file mode 100644 index b6a01eff2d..0000000000 --- a/src/armnn/layers/EqualLayer.hpp +++ /dev/null @@ -1,38 +0,0 @@ -// -// Copyright © 2017 Arm Ltd. All rights reserved. -// SPDX-License-Identifier: MIT -// - -#pragma once - -#include "ElementwiseBaseLayer.hpp" - -namespace armnn -{ -/// This layer represents an equal operation. -class EqualLayer : public ElementwiseBaseLayer -{ -public: - /// Makes a workload for the Equal type. - /// @param [in] graph The graph where this layer can be found. - /// @param [in] factory The workload factory which will create the workload. - /// @return A pointer to the created workload, or nullptr if not created. - virtual std::unique_ptr<IWorkload> CreateWorkload(const Graph& graph, - const IWorkloadFactory& factory) const override; - - /// Creates a dynamically-allocated copy of this layer. - /// @param [in] graph The graph into which this layer is being cloned. - EqualLayer* Clone(Graph& graph) const override; - - void Accept(ILayerVisitor& visitor) const override; - -protected: - /// Constructor to create a EqualLayer. - /// @param [in] name Optional name for the layer. - EqualLayer(const char* name); - - /// Default destructor - ~EqualLayer() = default; -}; - -} //namespace armnn diff --git a/src/armnn/layers/GreaterLayer.cpp b/src/armnn/layers/GreaterLayer.cpp deleted file mode 100644 index a9fe5e0d8c..0000000000 --- a/src/armnn/layers/GreaterLayer.cpp +++ /dev/null @@ -1,39 +0,0 @@ -// -// Copyright © 2017 Arm Ltd. All rights reserved. -// SPDX-License-Identifier: MIT -// - -#include "GreaterLayer.hpp" - -#include "LayerCloneBase.hpp" - -#include <armnn/TypesUtils.hpp> -#include <backendsCommon/WorkloadData.hpp> -#include <backendsCommon/WorkloadFactory.hpp> - -namespace armnn -{ - -GreaterLayer::GreaterLayer(const char* name) - : ElementwiseBaseLayer(2, 1, LayerType::Greater, name) -{ -} - -std::unique_ptr<IWorkload> GreaterLayer::CreateWorkload(const Graph& graph, - const IWorkloadFactory& factory) const -{ - GreaterQueueDescriptor descriptor; - return factory.CreateGreater(descriptor, PrepInfoAndDesc(descriptor, graph)); -} - -GreaterLayer* GreaterLayer::Clone(Graph& graph) const -{ - return CloneBase<GreaterLayer>(graph, GetName()); -} - -void GreaterLayer::Accept(ILayerVisitor& visitor) const -{ - visitor.VisitGreaterLayer(this, GetName()); -} - -} // namespace armnn diff --git a/src/armnn/layers/GreaterLayer.hpp b/src/armnn/layers/GreaterLayer.hpp deleted file mode 100644 index bdee948d6e..0000000000 --- a/src/armnn/layers/GreaterLayer.hpp +++ /dev/null @@ -1,39 +0,0 @@ -// -// Copyright © 2017 Arm Ltd. All rights reserved. -// SPDX-License-Identifier: MIT -// - -#pragma once - -#include "ElementwiseBaseLayer.hpp" - -namespace armnn -{ - -/// This layer represents a greater operation. -class GreaterLayer : public ElementwiseBaseLayer -{ -public: - /// Makes a workload for the Greater type. - /// @param [in] graph The graph where this layer can be found. - /// @param [in] factory The workload factory which will create the workload. - /// @return A pointer to the created workload, or nullptr if not created. - virtual std::unique_ptr<IWorkload> CreateWorkload(const Graph& graph, - const IWorkloadFactory& factory) const override; - - /// Creates a dynamically-allocated copy of this layer. - /// @param [in] graph The graph into which this layer is being cloned. - GreaterLayer* Clone(Graph& graph) const override; - - void Accept(ILayerVisitor& visitor) const override; - -protected: - /// Constructor to create a GreaterLayer. - /// @param [in] name Optional name for the layer. - GreaterLayer(const char* name); - - /// Default destructor - ~GreaterLayer() = default; -}; - -} //namespace armnn diff --git a/src/armnn/test/TestNameAndDescriptorLayerVisitor.cpp b/src/armnn/test/TestNameAndDescriptorLayerVisitor.cpp index 0b126235e8..36bbd36792 100644 --- a/src/armnn/test/TestNameAndDescriptorLayerVisitor.cpp +++ b/src/armnn/test/TestNameAndDescriptorLayerVisitor.cpp @@ -65,6 +65,12 @@ armnn::BatchToSpaceNdDescriptor GetDescriptor<armnn::BatchToSpaceNdDescriptor>() } template<> +armnn::ComparisonDescriptor GetDescriptor<armnn::ComparisonDescriptor>() +{ + return armnn::ComparisonDescriptor(armnn::ComparisonOperation::GreaterOrEqual); +} + +template<> armnn::ConcatDescriptor GetDescriptor<armnn::ConcatDescriptor>() { armnn::ConcatDescriptor descriptor(2, 2); @@ -243,6 +249,7 @@ TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(Activation) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(ArgMinMax) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(DepthToSpace) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(BatchToSpaceNd) +TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(Comparison) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(Concat) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(InstanceNormalization) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(L2Normalization) @@ -262,4 +269,4 @@ TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(Splitter) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(Stack) TEST_SUITE_NAME_AND_DESCRIPTOR_LAYER_VISITOR(StridedSlice) -BOOST_AUTO_TEST_SUITE_END()
\ No newline at end of file +BOOST_AUTO_TEST_SUITE_END() diff --git a/src/armnn/test/TestNameAndDescriptorLayerVisitor.hpp b/src/armnn/test/TestNameAndDescriptorLayerVisitor.hpp index aefcba5f59..b1f7f57075 100644 --- a/src/armnn/test/TestNameAndDescriptorLayerVisitor.hpp +++ b/src/armnn/test/TestNameAndDescriptorLayerVisitor.hpp @@ -46,6 +46,7 @@ public: \ DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(Activation) DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(ArgMinMax) DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(BatchToSpaceNd) +DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(Comparison) DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(Concat) DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(DepthToSpace) DECLARE_TEST_NAME_AND_DESCRIPTOR_LAYER_VISITOR_CLASS(InstanceNormalization) diff --git a/src/armnn/test/TestNameOnlyLayerVisitor.cpp b/src/armnn/test/TestNameOnlyLayerVisitor.cpp index 6bc2dc7c65..32de94e7ef 100644 --- a/src/armnn/test/TestNameOnlyLayerVisitor.cpp +++ b/src/armnn/test/TestNameOnlyLayerVisitor.cpp @@ -42,10 +42,8 @@ TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Abs) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Addition) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Dequantize) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Division) -TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Equal) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Floor) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Gather) -TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Greater) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Maximum) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Merge) TEST_SUITE_NAME_ONLY_LAYER_VISITOR(Minimum) diff --git a/src/armnn/test/TestNameOnlyLayerVisitor.hpp b/src/armnn/test/TestNameOnlyLayerVisitor.hpp index a772cb3283..c770b5e9e0 100644 --- a/src/armnn/test/TestNameOnlyLayerVisitor.hpp +++ b/src/armnn/test/TestNameOnlyLayerVisitor.hpp @@ -29,10 +29,8 @@ DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Abs) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Addition) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Dequantize) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Division) -DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Equal) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Floor) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Gather) -DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Greater) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Maximum) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Merge) DECLARE_TEST_NAME_ONLY_LAYER_VISITOR_CLASS(Minimum) diff --git a/src/armnnDeserializer/Deserializer.cpp b/src/armnnDeserializer/Deserializer.cpp index 67836c5843..0b18ccd051 100644 --- a/src/armnnDeserializer/Deserializer.cpp +++ b/src/armnnDeserializer/Deserializer.cpp @@ -1270,7 +1270,8 @@ void Deserializer::ParseEqual(GraphPtr graph, unsigned int layerIndex) CHECK_VALID_SIZE(outputs.size(), 1); auto layerName = GetLayerName(graph, layerIndex); - IConnectableLayer* layer = m_Network->AddEqualLayer(layerName.c_str()); + armnn::ComparisonDescriptor descriptor(armnn::ComparisonOperation::Equal); + IConnectableLayer* layer = m_Network->AddComparisonLayer(descriptor, layerName.c_str()); armnn::TensorInfo outputTensorInfo = ToTensorInfo(outputs[0]); layer->GetOutputSlot(0).SetTensorInfo(outputTensorInfo); @@ -1290,7 +1291,8 @@ void Deserializer::ParseGreater(GraphPtr graph, unsigned int layerIndex) CHECK_VALID_SIZE(outputs.size(), 1); auto layerName = GetLayerName(graph, layerIndex); - IConnectableLayer* layer = m_Network->AddGreaterLayer(layerName.c_str()); + armnn::ComparisonDescriptor descriptor(armnn::ComparisonOperation::Greater); + IConnectableLayer* layer = m_Network->AddComparisonLayer(descriptor, layerName.c_str()); armnn::TensorInfo outputTensorInfo = ToTensorInfo(outputs[0]); layer->GetOutputSlot(0).SetTensorInfo(outputTensorInfo); diff --git a/src/armnnSerializer/Serializer.cpp b/src/armnnSerializer/Serializer.cpp index 11f833c5b7..5c9855f87e 100644 --- a/src/armnnSerializer/Serializer.cpp +++ b/src/armnnSerializer/Serializer.cpp @@ -240,6 +240,13 @@ void SerializerVisitor::VisitBatchNormalizationLayer(const armnn::IConnectableLa CreateAnyLayer(fbBatchNormalizationLayer.o, serializer::Layer::Layer_BatchNormalizationLayer); } +void SerializerVisitor::VisitComparisonLayer(const armnn::IConnectableLayer* layer, + const armnn::ComparisonDescriptor& descriptor, + const char* name) +{ + throw armnn::UnimplementedException("SerializerVisitor::VisitComparisonLayer() is not implemented"); +} + // Build FlatBuffer for Constant Layer void SerializerVisitor::VisitConstantLayer(const armnn::IConnectableLayer* layer, const armnn::ConstTensor& input, diff --git a/src/armnnSerializer/Serializer.hpp b/src/armnnSerializer/Serializer.hpp index 8c13245aeb..79dc17ba01 100644 --- a/src/armnnSerializer/Serializer.hpp +++ b/src/armnnSerializer/Serializer.hpp @@ -68,6 +68,10 @@ public: const armnn::ConstTensor& gamma, const char* name = nullptr) override; + void VisitComparisonLayer(const armnn::IConnectableLayer* layer, + const armnn::ComparisonDescriptor& descriptor, + const char* name = nullptr) override; + void VisitConcatLayer(const armnn::IConnectableLayer* layer, const armnn::ConcatDescriptor& concatDescriptor, const char* name = nullptr) override; @@ -103,6 +107,7 @@ public: void VisitDivisionLayer(const armnn::IConnectableLayer* layer, const char* name = nullptr) override; + ARMNN_DEPRECATED_MSG("Use VisitComparisonLayer instead") void VisitEqualLayer(const armnn::IConnectableLayer* layer, const char* name = nullptr) override; @@ -118,6 +123,7 @@ public: void VisitGatherLayer(const armnn::IConnectableLayer* layer, const char* name = nullptr) override; + ARMNN_DEPRECATED_MSG("Use VisitComparisonLayer instead") void VisitGreaterLayer(const armnn::IConnectableLayer* layer, const char* name = nullptr) override; diff --git a/src/armnnSerializer/test/SerializerTests.cpp b/src/armnnSerializer/test/SerializerTests.cpp index a70c891849..58f56f484f 100644 --- a/src/armnnSerializer/test/SerializerTests.cpp +++ b/src/armnnSerializer/test/SerializerTests.cpp @@ -1009,48 +1009,6 @@ BOOST_AUTO_TEST_CASE(SerializeDivision) deserializedNetwork->Accept(verifier); } -BOOST_AUTO_TEST_CASE(SerializeEqual) -{ - class EqualLayerVerifier : public LayerVerifierBase - { - public: - EqualLayerVerifier(const std::string& layerName, - const std::vector<armnn::TensorInfo>& inputInfos, - const std::vector<armnn::TensorInfo>& outputInfos) - : LayerVerifierBase(layerName, inputInfos, outputInfos) {} - - void VisitEqualLayer(const armnn::IConnectableLayer* layer, const char* name) override - { - VerifyNameAndConnections(layer, name); - } - }; - - const std::string layerName("equal"); - const armnn::TensorInfo inputTensorInfo1 = armnn::TensorInfo({2, 1, 2, 4}, armnn::DataType::Float32); - const armnn::TensorInfo inputTensorInfo2 = armnn::TensorInfo({2, 1, 2, 4}, armnn::DataType::Float32); - const armnn::TensorInfo outputTensorInfo = armnn::TensorInfo({2, 1, 2, 4}, armnn::DataType::Boolean); - - armnn::INetworkPtr network = armnn::INetwork::Create(); - armnn::IConnectableLayer* const inputLayer1 = network->AddInputLayer(0); - armnn::IConnectableLayer* const inputLayer2 = network->AddInputLayer(1); - armnn::IConnectableLayer* const equalLayer = network->AddEqualLayer(layerName.c_str()); - armnn::IConnectableLayer* const outputLayer = network->AddOutputLayer(0); - - inputLayer1->GetOutputSlot(0).Connect(equalLayer->GetInputSlot(0)); - inputLayer2->GetOutputSlot(0).Connect(equalLayer->GetInputSlot(1)); - equalLayer->GetOutputSlot(0).Connect(outputLayer->GetInputSlot(0)); - - inputLayer1->GetOutputSlot(0).SetTensorInfo(inputTensorInfo1); - inputLayer2->GetOutputSlot(0).SetTensorInfo(inputTensorInfo2); - equalLayer->GetOutputSlot(0).SetTensorInfo(outputTensorInfo); - - armnn::INetworkPtr deserializedNetwork = DeserializeNetwork(SerializeNetwork(*network)); - BOOST_CHECK(deserializedNetwork); - - EqualLayerVerifier verifier(layerName, {inputTensorInfo1, inputTensorInfo2}, {outputTensorInfo}); - deserializedNetwork->Accept(verifier); -} - BOOST_AUTO_TEST_CASE(SerializeFloor) { class FloorLayerVerifier : public LayerVerifierBase @@ -1225,48 +1183,6 @@ BOOST_AUTO_TEST_CASE(SerializeGather) deserializedNetwork->Accept(verifier); } -BOOST_AUTO_TEST_CASE(SerializeGreater) -{ - class GreaterLayerVerifier : public LayerVerifierBase - { - public: - GreaterLayerVerifier(const std::string& layerName, - const std::vector<armnn::TensorInfo>& inputInfos, - const std::vector<armnn::TensorInfo>& outputInfos) - : LayerVerifierBase(layerName, inputInfos, outputInfos) {} - - void VisitGreaterLayer(const armnn::IConnectableLayer* layer, const char* name) override - { - VerifyNameAndConnections(layer, name); - } - }; - - const std::string layerName("greater"); - const armnn::TensorInfo inputTensorInfo1({ 1, 2, 2, 2 }, armnn::DataType::Float32); - const armnn::TensorInfo inputTensorInfo2({ 1, 2, 2, 2 }, armnn::DataType::Float32); - const armnn::TensorInfo outputTensorInfo({ 1, 2, 2, 2 }, armnn::DataType::Boolean); - - armnn::INetworkPtr network = armnn::INetwork::Create(); - armnn::IConnectableLayer* const inputLayer1 = network->AddInputLayer(0); - armnn::IConnectableLayer* const inputLayer2 = network->AddInputLayer(1); - armnn::IConnectableLayer* const greaterLayer = network->AddGreaterLayer(layerName.c_str()); - armnn::IConnectableLayer* const outputLayer = network->AddOutputLayer(0); - - inputLayer1->GetOutputSlot(0).Connect(greaterLayer->GetInputSlot(0)); - inputLayer2->GetOutputSlot(0).Connect(greaterLayer->GetInputSlot(1)); - greaterLayer->GetOutputSlot(0).Connect(outputLayer->GetInputSlot(0)); - - inputLayer1->GetOutputSlot(0).SetTensorInfo(inputTensorInfo1); - inputLayer2->GetOutputSlot(0).SetTensorInfo(inputTensorInfo2); - greaterLayer->GetOutputSlot(0).SetTensorInfo(outputTensorInfo); - - armnn::INetworkPtr deserializedNetwork = DeserializeNetwork(SerializeNetwork(*network)); - BOOST_CHECK(deserializedNetwork); - - GreaterLayerVerifier verifier(layerName, {inputTensorInfo1, inputTensorInfo2}, {outputTensorInfo}); - deserializedNetwork->Accept(verifier); -} - BOOST_AUTO_TEST_CASE(SerializeInstanceNormalization) { class InstanceNormalizationLayerVerifier : public LayerVerifierBase diff --git a/src/armnnTfParser/TfParser.cpp b/src/armnnTfParser/TfParser.cpp index 76d25d1d05..d085ed84e3 100755 --- a/src/armnnTfParser/TfParser.cpp +++ b/src/armnnTfParser/TfParser.cpp @@ -1862,7 +1862,8 @@ ParsedTfOperationPtr TfParser::ParseGreater(const tensorflow::NodeDef& nodeDef, IOutputSlot* input0Slot = inputLayers.first; IOutputSlot* input1Slot = inputLayers.second; - IConnectableLayer* const layer = m_Network->AddGreaterLayer(nodeDef.name().c_str()); + ComparisonDescriptor descriptor(ComparisonOperation::Greater); + IConnectableLayer* const layer = m_Network->AddComparisonLayer(descriptor, nodeDef.name().c_str()); return ProcessComparisonLayer(input0Slot, input1Slot, layer, nodeDef); } @@ -1874,7 +1875,8 @@ ParsedTfOperationPtr TfParser::ParseEqual(const tensorflow::NodeDef& nodeDef, IOutputSlot* input0Slot = inputLayers.first; IOutputSlot* input1Slot = inputLayers.second; - IConnectableLayer* const layer = m_Network->AddEqualLayer(nodeDef.name().c_str()); + ComparisonDescriptor descriptor(ComparisonOperation::Equal); + IConnectableLayer* const layer = m_Network->AddComparisonLayer(descriptor, nodeDef.name().c_str()); return ProcessComparisonLayer(input0Slot, input1Slot, layer, nodeDef); } diff --git a/src/backends/backendsCommon/LayerSupportBase.cpp b/src/backends/backendsCommon/LayerSupportBase.cpp index 7d5555ce68..358106e5e9 100644 --- a/src/backends/backendsCommon/LayerSupportBase.cpp +++ b/src/backends/backendsCommon/LayerSupportBase.cpp @@ -84,6 +84,15 @@ bool LayerSupportBase::IsBatchToSpaceNdSupported(const TensorInfo& input, return DefaultLayerSupport(__func__, __FILE__, __LINE__, reasonIfUnsupported); } +bool LayerSupportBase::IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported) const +{ + return DefaultLayerSupport(__func__, __FILE__, __LINE__, reasonIfUnsupported); +} + bool LayerSupportBase::IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const OriginsDescriptor& descriptor, diff --git a/src/backends/backendsCommon/LayerSupportBase.hpp b/src/backends/backendsCommon/LayerSupportBase.hpp index cb660f5c2b..d4c37c1a91 100644 --- a/src/backends/backendsCommon/LayerSupportBase.hpp +++ b/src/backends/backendsCommon/LayerSupportBase.hpp @@ -46,6 +46,12 @@ public: const BatchToSpaceNdDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const OriginsDescriptor& descriptor, @@ -108,6 +114,7 @@ public: const TensorInfo& output, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsEqualSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& output, @@ -133,6 +140,7 @@ public: const TensorInfo& output, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsGreaterSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& output, diff --git a/src/backends/backendsCommon/WorkloadData.cpp b/src/backends/backendsCommon/WorkloadData.cpp index b8d4f0dfff..cfb38b4820 100644 --- a/src/backends/backendsCommon/WorkloadData.cpp +++ b/src/backends/backendsCommon/WorkloadData.cpp @@ -2779,4 +2779,28 @@ void DepthToSpaceQueueDescriptor::Validate(const WorkloadInfo& workloadInfo) con } } +void ComparisonQueueDescriptor::Validate(const WorkloadInfo& workloadInfo) const +{ + const std::string descriptorName{"ComparisonQueueDescriptor"}; + + ValidateNumInputs(workloadInfo, descriptorName, 2); + ValidateNumOutputs(workloadInfo, descriptorName, 1); + + const TensorInfo& inputTensorInfo0 = workloadInfo.m_InputTensorInfos[0]; + const TensorInfo& inputTensorInfo1 = workloadInfo.m_InputTensorInfos[1]; + const TensorInfo& outputTensorInfo = workloadInfo.m_OutputTensorInfos[0]; + + ValidateBroadcastTensorShapesMatch(inputTensorInfo0, + inputTensorInfo1, + outputTensorInfo, + descriptorName, + "input_0", + "input_1"); + + if (outputTensorInfo.GetDataType() != DataType::Boolean) + { + throw InvalidArgumentException(descriptorName + ": Output tensor type must be Boolean."); + } +} + } // namespace armnn diff --git a/src/backends/backendsCommon/WorkloadData.hpp b/src/backends/backendsCommon/WorkloadData.hpp index 5a3600fc71..b45a1718f6 100644 --- a/src/backends/backendsCommon/WorkloadData.hpp +++ b/src/backends/backendsCommon/WorkloadData.hpp @@ -553,4 +553,9 @@ struct DepthToSpaceQueueDescriptor : QueueDescriptorWithParameters<DepthToSpaceD void Validate(const WorkloadInfo& workloadInfo) const; }; +struct ComparisonQueueDescriptor : QueueDescriptorWithParameters<ComparisonDescriptor> +{ + void Validate(const WorkloadInfo& workloadInfo) const; +}; + } // namespace armnn diff --git a/src/backends/backendsCommon/WorkloadFactory.cpp b/src/backends/backendsCommon/WorkloadFactory.cpp index f19b48491a..30dfa023f9 100644 --- a/src/backends/backendsCommon/WorkloadFactory.cpp +++ b/src/backends/backendsCommon/WorkloadFactory.cpp @@ -148,6 +148,21 @@ bool IWorkloadFactory::IsLayerSupported(const BackendId& backendId, reason); break; } + case LayerType::Comparison: + { + auto cLayer = boost::polymorphic_downcast<const ComparisonLayer*>(&layer); + + const TensorInfo& input0 = layer.GetInputSlot(0).GetConnection()->GetTensorInfo(); + const TensorInfo& input1 = layer.GetInputSlot(1).GetConnection()->GetTensorInfo(); + const TensorInfo& output = layer.GetOutputSlot(0).GetTensorInfo(); + + result = layerSupportObject->IsComparisonSupported(OverrideDataType(input0, dataType), + OverrideDataType(input1, dataType), + OverrideDataType(output, DataType::Boolean), + cLayer->GetParameters(), + reason); + break; + } case LayerType::Constant: { const TensorInfo& output = layer.GetOutputSlot(0).GetTensorInfo(); @@ -268,17 +283,6 @@ bool IWorkloadFactory::IsLayerSupported(const BackendId& backendId, reason); break; } - case LayerType::Equal: - { - const TensorInfo& input0 = layer.GetInputSlot(0).GetConnection()->GetTensorInfo(); - const TensorInfo& input1 = layer.GetInputSlot(1).GetConnection()->GetTensorInfo(); - const TensorInfo& output = layer.GetOutputSlot(0).GetTensorInfo(); - result = layerSupportObject->IsEqualSupported(OverrideDataType(input0, dataType), - OverrideDataType(input1, dataType), - OverrideDataType(output, dataType), - reason); - break; - } case LayerType::FakeQuantization: { auto cLayer = boost::polymorphic_downcast<const FakeQuantizationLayer*>(&layer); @@ -957,17 +961,6 @@ bool IWorkloadFactory::IsLayerSupported(const BackendId& backendId, reason); break; } - case LayerType::Greater: - { - const TensorInfo& input0 = layer.GetInputSlot(0).GetConnection()->GetTensorInfo(); - const TensorInfo& input1 = layer.GetInputSlot(1).GetConnection()->GetTensorInfo(); - const TensorInfo& output = layer.GetOutputSlot(0).GetTensorInfo(); - result = layerSupportObject->IsGreaterSupported(OverrideDataType(input0, dataType), - OverrideDataType(input1, dataType), - OverrideDataType(output, DataType::Boolean), - reason); - break; - } case LayerType::Prelu: { const TensorInfo& input = layer.GetInputSlot(0).GetConnection()->GetTensorInfo(); @@ -1065,6 +1058,12 @@ std::unique_ptr<IWorkload> IWorkloadFactory::CreateBatchToSpaceNd(const BatchToS return std::unique_ptr<IWorkload>(); } +std::unique_ptr<IWorkload> IWorkloadFactory::CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const +{ + return std::unique_ptr<IWorkload>(); +} + std::unique_ptr<IWorkload> IWorkloadFactory::CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const { diff --git a/src/backends/backendsCommon/WorkloadFactory.hpp b/src/backends/backendsCommon/WorkloadFactory.hpp index fa7a9d46a8..819b8c768b 100644 --- a/src/backends/backendsCommon/WorkloadFactory.hpp +++ b/src/backends/backendsCommon/WorkloadFactory.hpp @@ -69,6 +69,9 @@ public: virtual std::unique_ptr<IWorkload> CreateBatchToSpaceNd(const BatchToSpaceNdQueueDescriptor& descriptor, const WorkloadInfo& Info) const; + virtual std::unique_ptr<IWorkload> CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& Info) const; + virtual std::unique_ptr<IWorkload> CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const; @@ -102,6 +105,7 @@ public: virtual std::unique_ptr<IWorkload> CreateDivision(const DivisionQueueDescriptor& descriptor, const WorkloadInfo& info) const; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") virtual std::unique_ptr<IWorkload> CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& Info) const; @@ -117,6 +121,7 @@ public: virtual std::unique_ptr<IWorkload> CreateGather(const GatherQueueDescriptor& descriptor, const WorkloadInfo& info) const; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") virtual std::unique_ptr<IWorkload> CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const; diff --git a/src/backends/backendsCommon/test/ArithmeticTestImpl.hpp b/src/backends/backendsCommon/test/ArithmeticTestImpl.hpp deleted file mode 100644 index d0e85dd31d..0000000000 --- a/src/backends/backendsCommon/test/ArithmeticTestImpl.hpp +++ /dev/null @@ -1,113 +0,0 @@ -// -// Copyright © 2017 Arm Ltd. All rights reserved. -// SPDX-License-Identifier: MIT -// -#pragma once - -#include "CommonTestUtils.hpp" - -#include <ResolveType.hpp> - -#include <armnn/INetwork.hpp> - -#include <boost/test/unit_test.hpp> - -#include <vector> - -namespace -{ - -template<armnn::DataType ArmnnTypeInput, armnn::DataType ArmnnTypeOutput> -INetworkPtr CreateArithmeticNetwork(const std::vector<TensorShape>& inputShapes, - const TensorShape& outputShape, - const LayerType type, - const float qScale = 1.0f, - const int32_t qOffset = 0) -{ - using namespace armnn; - - // Builds up the structure of the network. - INetworkPtr net(INetwork::Create()); - - IConnectableLayer* arithmeticLayer = nullptr; - - switch(type){ - case LayerType::Equal: arithmeticLayer = net->AddEqualLayer("equal"); break; - case LayerType::Greater: arithmeticLayer = net->AddGreaterLayer("greater"); break; - default: BOOST_TEST_FAIL("Non-Arithmetic layer type called."); - } - - for (unsigned int i = 0; i < inputShapes.size(); ++i) - { - TensorInfo inputTensorInfo(inputShapes[i], ArmnnTypeInput, qScale, qOffset); - IConnectableLayer* input = net->AddInputLayer(boost::numeric_cast<LayerBindingId>(i)); - Connect(input, arithmeticLayer, inputTensorInfo, 0, i); - } - - TensorInfo outputTensorInfo(outputShape, ArmnnTypeOutput, qScale, qOffset); - IConnectableLayer* output = net->AddOutputLayer(0, "output"); - Connect(arithmeticLayer, output, outputTensorInfo, 0, 0); - - return net; -} - -template<armnn::DataType ArmnnInputType, - armnn::DataType ArmnnOutputType, - typename TInput = armnn::ResolveType<ArmnnInputType>, - typename TOutput = armnn::ResolveType<ArmnnOutputType>> -void ArithmeticSimpleEndToEnd(const std::vector<BackendId>& backends, - const LayerType type, - const std::vector<TOutput> expectedOutput) -{ - using namespace armnn; - - const std::vector<TensorShape> inputShapes{{ 2, 2, 2, 2 }, { 2, 2, 2, 2 }}; - const TensorShape& outputShape = { 2, 2, 2, 2 }; - - // Builds up the structure of the network - INetworkPtr net = CreateArithmeticNetwork<ArmnnInputType, ArmnnOutputType>(inputShapes, outputShape, type); - - BOOST_TEST_CHECKPOINT("create a network"); - - const std::vector<TInput> input0({ 1, 1, 1, 1, 5, 5, 5, 5, - 3, 3, 3, 3, 4, 4, 4, 4 }); - - const std::vector<TInput> input1({ 1, 1, 1, 1, 3, 3, 3, 3, - 5, 5, 5, 5, 4, 4, 4, 4 }); - - std::map<int, std::vector<TInput>> inputTensorData = {{ 0, input0 }, { 1, input1 }}; - std::map<int, std::vector<TOutput>> expectedOutputData = {{ 0, expectedOutput }}; - - EndToEndLayerTestImpl<ArmnnInputType, ArmnnOutputType>(move(net), inputTensorData, expectedOutputData, backends); -} - -template<armnn::DataType ArmnnInputType, - armnn::DataType ArmnnOutputType, - typename TInput = armnn::ResolveType<ArmnnInputType>, - typename TOutput = armnn::ResolveType<ArmnnOutputType>> -void ArithmeticBroadcastEndToEnd(const std::vector<BackendId>& backends, - const LayerType type, - const std::vector<TOutput> expectedOutput) -{ - using namespace armnn; - - const std::vector<TensorShape> inputShapes{{ 1, 2, 2, 3 }, { 1, 1, 1, 3 }}; - const TensorShape& outputShape = { 1, 2, 2, 3 }; - - // Builds up the structure of the network - INetworkPtr net = CreateArithmeticNetwork<ArmnnInputType, ArmnnOutputType>(inputShapes, outputShape, type); - - BOOST_TEST_CHECKPOINT("create a network"); - - const std::vector<TInput> input0({ 1, 2, 3, 1, 0, 6, - 7, 8, 9, 10, 11, 12 }); - - const std::vector<TInput> input1({ 1, 1, 3 }); - - std::map<int, std::vector<TInput>> inputTensorData = {{ 0, input0 }, { 1, input1 }}; - std::map<int, std::vector<TOutput>> expectedOutputData = {{ 0, expectedOutput }}; - - EndToEndLayerTestImpl<ArmnnInputType, ArmnnOutputType>(move(net), inputTensorData, expectedOutputData, backends); -} - -} // anonymous namespace diff --git a/src/backends/backendsCommon/test/CMakeLists.txt b/src/backends/backendsCommon/test/CMakeLists.txt index 797dc90952..7449e69dcd 100644 --- a/src/backends/backendsCommon/test/CMakeLists.txt +++ b/src/backends/backendsCommon/test/CMakeLists.txt @@ -10,6 +10,7 @@ list(APPEND armnnBackendsCommonUnitTests_sources BackendRegistryTests.cpp CommonTestUtils.cpp CommonTestUtils.hpp + ComparisonEndToEndTestImpl.hpp DataLayoutUtils.hpp DataTypeUtils.hpp DepthToSpaceEndToEndTestImpl.hpp @@ -58,6 +59,7 @@ list(APPEND armnnBackendsCommonUnitTests_sources layerTests/BatchNormalizationTestImpl.cpp layerTests/BatchNormalizationTestImpl.hpp layerTests/BatchToSpaceNdTestImpl.hpp + layerTests/ComparisonTestImpl.hpp layerTests/ConcatTestImpl.cpp layerTests/ConcatTestImpl.hpp layerTests/ConstantTestImpl.cpp diff --git a/src/backends/backendsCommon/test/ComparisonEndToEndTestImpl.hpp b/src/backends/backendsCommon/test/ComparisonEndToEndTestImpl.hpp new file mode 100644 index 0000000000..dc53b7b246 --- /dev/null +++ b/src/backends/backendsCommon/test/ComparisonEndToEndTestImpl.hpp @@ -0,0 +1,103 @@ +// +// Copyright © 2017 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// +#pragma once + +#include "CommonTestUtils.hpp" + +#include <ResolveType.hpp> + +#include <armnn/INetwork.hpp> + +#include <boost/test/unit_test.hpp> + +#include <vector> + +namespace +{ + +template<armnn::DataType ArmnnTypeInput> +INetworkPtr CreateComparisonNetwork(const std::vector<TensorShape>& inputShapes, + const TensorShape& outputShape, + ComparisonOperation operation, + const float qScale = 1.0f, + const int32_t qOffset = 0) +{ + using namespace armnn; + + INetworkPtr net(INetwork::Create()); + + ComparisonDescriptor descriptor(operation); + IConnectableLayer* comparisonLayer = net->AddComparisonLayer(descriptor, "comparison"); + + for (unsigned int i = 0; i < inputShapes.size(); ++i) + { + TensorInfo inputTensorInfo(inputShapes[i], ArmnnTypeInput, qScale, qOffset); + IConnectableLayer* input = net->AddInputLayer(boost::numeric_cast<LayerBindingId>(i)); + Connect(input, comparisonLayer, inputTensorInfo, 0, i); + } + + TensorInfo outputTensorInfo(outputShape, DataType::Boolean, qScale, qOffset); + IConnectableLayer* output = net->AddOutputLayer(0, "output"); + Connect(comparisonLayer, output, outputTensorInfo, 0, 0); + + return net; +} + +template<armnn::DataType ArmnnInType, + typename TInput = armnn::ResolveType<ArmnnInType>> +void ComparisonSimpleEndToEnd(const std::vector<BackendId>& backends, + ComparisonOperation operation, + const std::vector<uint8_t> expectedOutput) +{ + using namespace armnn; + + const std::vector<TensorShape> inputShapes{{ 2, 2, 2, 2 }, { 2, 2, 2, 2 }}; + const TensorShape& outputShape = { 2, 2, 2, 2 }; + + // Builds up the structure of the network + INetworkPtr net = CreateComparisonNetwork<ArmnnInType>(inputShapes, outputShape, operation); + + BOOST_TEST_CHECKPOINT("create a network"); + + const std::vector<TInput> input0({ 1, 1, 1, 1, 5, 5, 5, 5, + 3, 3, 3, 3, 4, 4, 4, 4 }); + + const std::vector<TInput> input1({ 1, 1, 1, 1, 3, 3, 3, 3, + 5, 5, 5, 5, 4, 4, 4, 4 }); + + std::map<int, std::vector<TInput>> inputTensorData = {{ 0, input0 }, { 1, input1 }}; + std::map<int, std::vector<uint8_t>> expectedOutputData = {{ 0, expectedOutput }}; + + EndToEndLayerTestImpl<ArmnnInType, DataType::Boolean>(move(net), inputTensorData, expectedOutputData, backends); +} + +template<armnn::DataType ArmnnInType, + typename TInput = armnn::ResolveType<ArmnnInType>> +void ComparisonBroadcastEndToEnd(const std::vector<BackendId>& backends, + ComparisonOperation operation, + const std::vector<uint8_t> expectedOutput) +{ + using namespace armnn; + + const std::vector<TensorShape> inputShapes{{ 1, 2, 2, 3 }, { 1, 1, 1, 3 }}; + const TensorShape& outputShape = { 1, 2, 2, 3 }; + + // Builds up the structure of the network + INetworkPtr net = CreateComparisonNetwork<ArmnnInType>(inputShapes, outputShape, operation); + + BOOST_TEST_CHECKPOINT("create a network"); + + const std::vector<TInput> input0({ 1, 2, 3, 1, 0, 6, + 7, 8, 9, 10, 11, 12 }); + + const std::vector<TInput> input1({ 1, 1, 3 }); + + std::map<int, std::vector<TInput>> inputTensorData = {{ 0, input0 }, { 1, input1 }}; + std::map<int, std::vector<uint8_t>> expectedOutputData = {{ 0, expectedOutput }}; + + EndToEndLayerTestImpl<ArmnnInType, DataType::Boolean>(move(net), inputTensorData, expectedOutputData, backends); +} + +} // anonymous namespace diff --git a/src/backends/backendsCommon/test/IsLayerSupportedTestImpl.hpp b/src/backends/backendsCommon/test/IsLayerSupportedTestImpl.hpp index 907285c5cf..9bddae9759 100644 --- a/src/backends/backendsCommon/test/IsLayerSupportedTestImpl.hpp +++ b/src/backends/backendsCommon/test/IsLayerSupportedTestImpl.hpp @@ -397,6 +397,8 @@ DECLARE_LAYER_POLICY_2_PARAM(BatchNormalization) DECLARE_LAYER_POLICY_2_PARAM(BatchToSpaceNd) +DECLARE_LAYER_POLICY_2_PARAM(Comparison) + DECLARE_LAYER_POLICY_2_PARAM(Concat) DECLARE_LAYER_POLICY_1_PARAM(Constant) @@ -421,8 +423,6 @@ DECLARE_LAYER_POLICY_1_PARAM(Dequantize) DECLARE_LAYER_POLICY_2_PARAM(DetectionPostProcess) -DECLARE_LAYER_POLICY_1_PARAM(Equal) - DECLARE_LAYER_POLICY_2_PARAM(FakeQuantization) DECLARE_LAYER_POLICY_1_PARAM(Floor) @@ -431,8 +431,6 @@ DECLARE_LAYER_POLICY_2_PARAM(FullyConnected) DECLARE_LAYER_POLICY_1_PARAM(Gather) -DECLARE_LAYER_POLICY_1_PARAM(Greater) - DECLARE_LAYER_POLICY_CUSTOM_PARAM(Input, armnn::LayerBindingId) DECLARE_LAYER_POLICY_2_PARAM(InstanceNormalization) diff --git a/src/backends/backendsCommon/test/layerTests/ComparisonTestImpl.hpp b/src/backends/backendsCommon/test/layerTests/ComparisonTestImpl.hpp new file mode 100644 index 0000000000..6ce9b306c0 --- /dev/null +++ b/src/backends/backendsCommon/test/layerTests/ComparisonTestImpl.hpp @@ -0,0 +1,126 @@ +// +// Copyright © 2019 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// + +#pragma once + +#include "LayerTestResult.hpp" + +#include <armnn/ArmNN.hpp> + +#include <ResolveType.hpp> + +#include <backendsCommon/IBackendInternal.hpp> +#include <backendsCommon/Workload.hpp> +#include <backendsCommon/WorkloadData.hpp> +#include <backendsCommon/WorkloadFactory.hpp> + +#include <backendsCommon/test/TensorCopyUtils.hpp> +#include <backendsCommon/test/WorkloadTestUtils.hpp> + +#include <test/TensorHelpers.hpp> + +template <std::size_t NumDims, + armnn::DataType ArmnnInType, + typename InType = armnn::ResolveType<ArmnnInType>> +LayerTestResult<uint8_t, NumDims> ComparisonTestImpl( + armnn::IWorkloadFactory & workloadFactory, + const armnn::IBackendInternal::IMemoryManagerSharedPtr & memoryManager, + const armnn::ComparisonDescriptor& descriptor, + const unsigned int shape0[NumDims], + std::vector<InType> values0, + float quantScale0, + int quantOffset0, + const unsigned int shape1[NumDims], + std::vector<InType> values1, + float quantScale1, + int quantOffset1, + const unsigned int outShape[NumDims], + std::vector<uint8_t> outValues, + float outQuantScale, + int outQuantOffset) +{ + armnn::TensorInfo inputTensorInfo0{NumDims, shape0, ArmnnInType}; + armnn::TensorInfo inputTensorInfo1{NumDims, shape1, ArmnnInType}; + armnn::TensorInfo outputTensorInfo{NumDims, outShape, armnn::DataType::Boolean}; + + auto input0 = MakeTensor<InType, NumDims>(inputTensorInfo0, values0); + auto input1 = MakeTensor<InType, NumDims>(inputTensorInfo1, values1); + + inputTensorInfo0.SetQuantizationScale(quantScale0); + inputTensorInfo0.SetQuantizationOffset(quantOffset0); + + inputTensorInfo1.SetQuantizationScale(quantScale1); + inputTensorInfo1.SetQuantizationOffset(quantOffset1); + + outputTensorInfo.SetQuantizationScale(outQuantScale); + outputTensorInfo.SetQuantizationOffset(outQuantOffset); + + LayerTestResult<uint8_t, NumDims> ret(outputTensorInfo); + + std::unique_ptr<armnn::ITensorHandle> inputHandle0 = workloadFactory.CreateTensorHandle(inputTensorInfo0); + std::unique_ptr<armnn::ITensorHandle> inputHandle1 = workloadFactory.CreateTensorHandle(inputTensorInfo1); + std::unique_ptr<armnn::ITensorHandle> outputHandle = workloadFactory.CreateTensorHandle(outputTensorInfo); + + armnn::ComparisonQueueDescriptor qDescriptor; + qDescriptor.m_Parameters = descriptor; + + armnn::WorkloadInfo info; + AddInputToWorkload(qDescriptor, info, inputTensorInfo0, inputHandle0.get()); + AddInputToWorkload(qDescriptor, info, inputTensorInfo1, inputHandle1.get()); + AddOutputToWorkload(qDescriptor, info, outputTensorInfo, outputHandle.get()); + + std::unique_ptr<armnn::IWorkload> workload = workloadFactory.CreateComparison(qDescriptor, info); + + inputHandle0->Allocate(); + inputHandle1->Allocate(); + outputHandle->Allocate(); + + CopyDataToITensorHandle(inputHandle0.get(), input0.origin()); + CopyDataToITensorHandle(inputHandle1.get(), input1.origin()); + + workload->PostAllocationConfigure(); + ExecuteWorkload(*workload, memoryManager); + + CopyDataFromITensorHandle(ret.output.origin(), outputHandle.get()); + + ret.outputExpected = MakeTensor<uint8_t, NumDims>(outputTensorInfo, outValues); + ret.compareBoolean = true; + + return ret; +} + +template <std::size_t NumDims, + armnn::DataType ArmnnInType, + typename InType = armnn::ResolveType<ArmnnInType>> +LayerTestResult<uint8_t, NumDims> ComparisonTestImpl( + armnn::IWorkloadFactory & workloadFactory, + const armnn::IBackendInternal::IMemoryManagerSharedPtr & memoryManager, + const armnn::ComparisonDescriptor& descriptor, + const unsigned int shape0[NumDims], + std::vector<InType> values0, + const unsigned int shape1[NumDims], + std::vector<InType> values1, + const unsigned int outShape[NumDims], + std::vector<uint8_t> outValues, + float qScale = 10.f, + int qOffset = 0) +{ + return ComparisonTestImpl<NumDims, ArmnnInType>( + workloadFactory, + memoryManager, + descriptor, + shape0, + values0, + qScale, + qOffset, + shape1, + values1, + qScale, + qOffset, + outShape, + outValues, + qScale, + qOffset); +} diff --git a/src/backends/backendsCommon/test/layerTests/EqualTestImpl.cpp b/src/backends/backendsCommon/test/layerTests/EqualTestImpl.cpp index b0b613c137..a3d2b2796f 100644 --- a/src/backends/backendsCommon/test/layerTests/EqualTestImpl.cpp +++ b/src/backends/backendsCommon/test/layerTests/EqualTestImpl.cpp @@ -4,18 +4,10 @@ // #include "EqualTestImpl.hpp" -#include "ElementwiseTestImpl.hpp" -#include <Half.hpp> +#include "ComparisonTestImpl.hpp" -template<> -std::unique_ptr<armnn::IWorkload> CreateWorkload<armnn::EqualQueueDescriptor>( - const armnn::IWorkloadFactory& workloadFactory, - const armnn::WorkloadInfo& info, - const armnn::EqualQueueDescriptor& descriptor) -{ - return workloadFactory.CreateEqual(descriptor, info); -} +#include <Half.hpp> LayerTestResult<uint8_t, 4> EqualSimpleTest(armnn::IWorkloadFactory& workloadFactory, const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager) @@ -39,9 +31,10 @@ LayerTestResult<uint8_t, 4> EqualSimpleTest(armnn::IWorkloadFactory& workloadFac std::vector<uint8_t> output({ 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1 }); - return ElementwiseTestHelper<4, armnn::EqualQueueDescriptor, armnn::DataType::Float32, armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape, input0, shape, @@ -62,9 +55,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1ElementTest( std::vector<uint8_t> output({ 1, 0, 0, 0, 0, 0, 0, 0}); - return ElementwiseTestHelper<4, armnn::EqualQueueDescriptor, armnn::DataType::Float32, armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, @@ -88,9 +82,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1DVectorTest( std::vector<uint8_t> output({ 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, armnn::EqualQueueDescriptor, armnn::DataType::Float32, armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, @@ -117,12 +112,10 @@ LayerTestResult<uint8_t, 4> EqualFloat16Test( std::vector<uint8_t> output({ 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape, input0, shape, @@ -148,12 +141,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1ElementFloat16Test( std::vector<uint8_t> output({ 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, @@ -179,12 +170,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1DVectorFloat16Test( std::vector<uint8_t> output({ 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, @@ -209,12 +198,10 @@ LayerTestResult<uint8_t, 4> EqualUint8Test( std::vector<uint8_t> output({ 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape, input0, shape, @@ -238,12 +225,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1ElementUint8Test( std::vector<uint8_t> output({ 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, @@ -267,12 +252,10 @@ LayerTestResult<uint8_t, 4> EqualBroadcast1DVectorUint8Test( std::vector<uint8_t> output({ 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0 }); - return ElementwiseTestHelper<4, - armnn::EqualQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Equal), shape0, input0, shape1, diff --git a/src/backends/backendsCommon/test/layerTests/GreaterTestImpl.cpp b/src/backends/backendsCommon/test/layerTests/GreaterTestImpl.cpp index 0148216285..271bc235a9 100644 --- a/src/backends/backendsCommon/test/layerTests/GreaterTestImpl.cpp +++ b/src/backends/backendsCommon/test/layerTests/GreaterTestImpl.cpp @@ -4,18 +4,10 @@ // #include "GreaterTestImpl.hpp" -#include "ElementwiseTestImpl.hpp" -#include <Half.hpp> +#include "ComparisonTestImpl.hpp" -template<> -std::unique_ptr<armnn::IWorkload> CreateWorkload<armnn::GreaterQueueDescriptor>( - const armnn::IWorkloadFactory& workloadFactory, - const armnn::WorkloadInfo& info, - const armnn::GreaterQueueDescriptor& descriptor) -{ - return workloadFactory.CreateGreater(descriptor, info); -} +#include <Half.hpp> LayerTestResult<uint8_t, 4> GreaterSimpleTest(armnn::IWorkloadFactory& workloadFactory, const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager) @@ -45,12 +37,10 @@ LayerTestResult<uint8_t, 4> GreaterSimpleTest(armnn::IWorkloadFactory& workloadF 0, 0, 0, 0, 0, 0, 0, 0 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float32, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape, input0, shape, @@ -71,12 +61,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1ElementTest( std::vector<uint8_t> output = { 0, 1, 1, 1, 1, 1, 1, 1}; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float32, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, @@ -106,12 +94,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1DVectorTest( 1, 1, 1, 1, 1, 1 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float32, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float32>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, @@ -151,12 +137,10 @@ LayerTestResult<uint8_t, 4> GreaterFloat16Test( 0, 0, 0, 0, 0, 0, 0, 0 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4,armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape, input0, shape, @@ -179,12 +163,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1ElementFloat16Test( std::vector<uint8_t> output = { 0, 1, 1, 1, 1, 1, 1, 1}; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, @@ -198,7 +180,7 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1DVectorFloat16Test( const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager) { using namespace half_float::literal; - + const unsigned int shape0[] = { 1, 2, 2, 3 }; const unsigned int shape1[] = { 1, 1, 1, 3 }; @@ -216,12 +198,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1DVectorFloat16Test( 1, 1, 1, 1, 1, 1 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::Float16, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::Float16>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, @@ -255,12 +235,10 @@ LayerTestResult<uint8_t, 4> GreaterUint8Test( 1, 1, 1, 1, 0, 0, 0, 0 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape, input0, shape, @@ -290,12 +268,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1ElementUint8Test( 1, 1, 1, 1, 1, 1 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, @@ -325,12 +301,10 @@ LayerTestResult<uint8_t, 4> GreaterBroadcast1DVectorUint8Test( 1, 1, 1, 1, 1, 1 }; - return ElementwiseTestHelper<4, - armnn::GreaterQueueDescriptor, - armnn::DataType::QuantisedAsymm8, - armnn::DataType::Boolean>( + return ComparisonTestImpl<4, armnn::DataType::QuantisedAsymm8>( workloadFactory, memoryManager, + armnn::ComparisonDescriptor(armnn::ComparisonOperation::Greater), shape0, input0, shape1, diff --git a/src/backends/cl/ClLayerSupport.cpp b/src/backends/cl/ClLayerSupport.cpp index c5ed8bff2a..bd2be57386 100644 --- a/src/backends/cl/ClLayerSupport.cpp +++ b/src/backends/cl/ClLayerSupport.cpp @@ -209,6 +209,24 @@ bool ClLayerSupport::IsBatchToSpaceNdSupported(const TensorInfo& input, descriptor); } +bool ClLayerSupport::IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported) const +{ + if (descriptor.m_Operation == ComparisonOperation::Greater) + { + FORWARD_WORKLOAD_VALIDATE_FUNC(ClGreaterWorkloadValidate, + reasonIfUnsupported, + input0, + input1, + output); + } + + return false; +} + bool ClLayerSupport::IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -398,11 +416,8 @@ bool ClLayerSupport::IsGreaterSupported(const TensorInfo& input0, const TensorInfo& output, Optional<std::string&> reasonIfUnsupported) const { - FORWARD_WORKLOAD_VALIDATE_FUNC(ClGreaterWorkloadValidate, - reasonIfUnsupported, - input0, - input1, - output); + ComparisonDescriptor descriptor(ComparisonOperation::Greater); + return IsComparisonSupported(input0, input1, output, descriptor, reasonIfUnsupported); } bool ClLayerSupport::IsInputSupported(const TensorInfo& input, diff --git a/src/backends/cl/ClLayerSupport.hpp b/src/backends/cl/ClLayerSupport.hpp index 59e849316f..26eb42e092 100644 --- a/src/backends/cl/ClLayerSupport.hpp +++ b/src/backends/cl/ClLayerSupport.hpp @@ -40,6 +40,12 @@ public: const BatchToSpaceNdDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& ouput, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -102,6 +108,7 @@ public: const FullyConnectedDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsGreaterSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& ouput, diff --git a/src/backends/cl/ClWorkloadFactory.cpp b/src/backends/cl/ClWorkloadFactory.cpp index c427ae7e12..04e09f4ff1 100644 --- a/src/backends/cl/ClWorkloadFactory.cpp +++ b/src/backends/cl/ClWorkloadFactory.cpp @@ -157,6 +157,20 @@ std::unique_ptr<IWorkload> ClWorkloadFactory::CreateBatchToSpaceNd(const BatchTo return MakeWorkload<ClBatchToSpaceNdWorkload>(descriptor, info); } +std::unique_ptr<IWorkload> ClWorkloadFactory::CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const +{ + if (descriptor.m_Parameters.m_Operation == ComparisonOperation::Greater) + { + GreaterQueueDescriptor greaterQueueDescriptor; + greaterQueueDescriptor.m_Inputs = descriptor.m_Inputs; + greaterQueueDescriptor.m_Outputs = descriptor.m_Outputs; + + return MakeWorkload<ClGreaterFloat32Workload, ClGreaterUint8Workload>(greaterQueueDescriptor, info); + } + return MakeWorkload<NullWorkload, NullWorkload>(descriptor, info); +} + std::unique_ptr<IWorkload> ClWorkloadFactory::CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const { @@ -230,7 +244,12 @@ std::unique_ptr<IWorkload> ClWorkloadFactory::CreateDivision(const DivisionQueue std::unique_ptr<IWorkload> ClWorkloadFactory::CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return MakeWorkload<NullWorkload, NullWorkload>(descriptor, info); + boost::ignore_unused(descriptor); + + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters = ComparisonDescriptor(ComparisonOperation::Equal); + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> ClWorkloadFactory::CreateFakeQuantization( @@ -261,7 +280,12 @@ std::unique_ptr<IWorkload> ClWorkloadFactory::CreateGather(const GatherQueueDesc std::unique_ptr<IWorkload> ClWorkloadFactory::CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return MakeWorkload<ClGreaterFloat32Workload, ClGreaterUint8Workload>(descriptor, info); + boost::ignore_unused(descriptor); + + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters = ComparisonDescriptor(ComparisonOperation::Greater); + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> ClWorkloadFactory::CreateInput(const InputQueueDescriptor& descriptor, diff --git a/src/backends/cl/ClWorkloadFactory.hpp b/src/backends/cl/ClWorkloadFactory.hpp index 9dbc615a4e..1cae6e1faf 100644 --- a/src/backends/cl/ClWorkloadFactory.hpp +++ b/src/backends/cl/ClWorkloadFactory.hpp @@ -53,6 +53,9 @@ public: std::unique_ptr<IWorkload> CreateBatchToSpaceNd(const BatchToSpaceNdQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + std::unique_ptr<IWorkload> CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const override; + std::unique_ptr<IWorkload> CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const override; diff --git a/src/backends/cl/test/ClEndToEndTests.cpp b/src/backends/cl/test/ClEndToEndTests.cpp index 59d26edf22..26f15b77da 100644 --- a/src/backends/cl/test/ClEndToEndTests.cpp +++ b/src/backends/cl/test/ClEndToEndTests.cpp @@ -6,7 +6,7 @@ #include <backendsCommon/test/EndToEndTestImpl.hpp> #include <backendsCommon/test/AbsEndToEndTestImpl.hpp> -#include <backendsCommon/test/ArithmeticTestImpl.hpp> +#include <backendsCommon/test/ComparisonEndToEndTestImpl.hpp> #include <backendsCommon/test/ConcatEndToEndTestImpl.hpp> #include <backendsCommon/test/DepthToSpaceEndToEndTestImpl.hpp> #include <backendsCommon/test/DequantizeEndToEndTestImpl.hpp> @@ -122,9 +122,9 @@ BOOST_AUTO_TEST_CASE(ClGreaterSimpleEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(ClGreaterSimpleEndToEndUint8Test) @@ -132,9 +132,9 @@ BOOST_AUTO_TEST_CASE(ClGreaterSimpleEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(ClGreaterBroadcastEndToEndTest) @@ -142,9 +142,9 @@ BOOST_AUTO_TEST_CASE(ClGreaterBroadcastEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(ClGreaterBroadcastEndToEndUint8Test) @@ -152,9 +152,9 @@ BOOST_AUTO_TEST_CASE(ClGreaterBroadcastEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } // InstanceNormalization diff --git a/src/backends/neon/NeonLayerSupport.cpp b/src/backends/neon/NeonLayerSupport.cpp index 270cb6264f..cc96f63c1a 100644 --- a/src/backends/neon/NeonLayerSupport.cpp +++ b/src/backends/neon/NeonLayerSupport.cpp @@ -180,6 +180,24 @@ bool NeonLayerSupport::IsBatchNormalizationSupported(const TensorInfo& input, descriptor); } +bool NeonLayerSupport::IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported) const +{ + if (descriptor.m_Operation == ComparisonOperation::Greater) + { + FORWARD_WORKLOAD_VALIDATE_FUNC(NeonGreaterWorkloadValidate, + reasonIfUnsupported, + input0, + input1, + output); + } + + return false; +} + bool NeonLayerSupport::IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -354,11 +372,8 @@ bool NeonLayerSupport::IsGreaterSupported(const armnn::TensorInfo& input0, const armnn::TensorInfo& output, armnn::Optional<std::string&> reasonIfUnsupported) const { - FORWARD_WORKLOAD_VALIDATE_FUNC(NeonGreaterWorkloadValidate, - reasonIfUnsupported, - input0, - input1, - output); + ComparisonDescriptor descriptor(ComparisonOperation::Greater); + return IsComparisonSupported(input0, input1, output, descriptor, reasonIfUnsupported); } bool NeonLayerSupport::IsInputSupported(const TensorInfo& input, diff --git a/src/backends/neon/NeonLayerSupport.hpp b/src/backends/neon/NeonLayerSupport.hpp index d6a24ad43b..9c1ca2a729 100644 --- a/src/backends/neon/NeonLayerSupport.hpp +++ b/src/backends/neon/NeonLayerSupport.hpp @@ -40,6 +40,12 @@ public: const BatchNormalizationDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -98,6 +104,7 @@ public: const FullyConnectedDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsGreaterSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& output, diff --git a/src/backends/neon/NeonWorkloadFactory.cpp b/src/backends/neon/NeonWorkloadFactory.cpp index 5bd8f293c5..dda1d7a132 100644 --- a/src/backends/neon/NeonWorkloadFactory.cpp +++ b/src/backends/neon/NeonWorkloadFactory.cpp @@ -131,6 +131,20 @@ std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateBatchToSpaceNd(const Batch return MakeWorkloadHelper<NullWorkload, NullWorkload>(descriptor, info); } +std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const +{ + if (descriptor.m_Parameters.m_Operation == ComparisonOperation::Greater) + { + GreaterQueueDescriptor greaterQueueDescriptor; + greaterQueueDescriptor.m_Inputs = descriptor.m_Inputs; + greaterQueueDescriptor.m_Outputs = descriptor.m_Outputs; + + return MakeWorkloadHelper<NeonGreaterFloat32Workload, NeonGreaterUint8Workload>(greaterQueueDescriptor, info); + } + return MakeWorkloadHelper<NullWorkload, NullWorkload>(descriptor, info); +} + std::unique_ptr<armnn::IWorkload> NeonWorkloadFactory::CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const { @@ -203,7 +217,12 @@ std::unique_ptr<armnn::IWorkload> NeonWorkloadFactory::CreateDivision( std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return MakeWorkloadHelper<NullWorkload, NullWorkload>(descriptor, info); + boost::ignore_unused(descriptor); + + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters = ComparisonDescriptor(ComparisonOperation::Equal); + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateFakeQuantization( @@ -235,7 +254,12 @@ std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateGather(const armnn::Gather std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return MakeWorkloadHelper<NeonGreaterFloat32Workload, NeonGreaterUint8Workload>(descriptor, info); + boost::ignore_unused(descriptor); + + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters = ComparisonDescriptor(ComparisonOperation::Greater); + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateInput(const InputQueueDescriptor& descriptor, diff --git a/src/backends/neon/NeonWorkloadFactory.hpp b/src/backends/neon/NeonWorkloadFactory.hpp index 9546164d36..5bee771528 100644 --- a/src/backends/neon/NeonWorkloadFactory.hpp +++ b/src/backends/neon/NeonWorkloadFactory.hpp @@ -57,6 +57,9 @@ public: std::unique_ptr<IWorkload> CreateBatchToSpaceNd(const BatchToSpaceNdQueueDescriptor& descriptor, const WorkloadInfo& Info) const override; + std::unique_ptr<IWorkload> CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& Info) const override; + std::unique_ptr<IWorkload> CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const override; @@ -90,6 +93,7 @@ public: std::unique_ptr<IWorkload> CreateDivision(const DivisionQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") std::unique_ptr<IWorkload> CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& info) const override; @@ -105,6 +109,7 @@ public: std::unique_ptr<IWorkload> CreateGather(const GatherQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") std::unique_ptr<IWorkload> CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const override; diff --git a/src/backends/neon/test/NeonEndToEndTests.cpp b/src/backends/neon/test/NeonEndToEndTests.cpp index 88f7ae7b5d..5146a598c7 100644 --- a/src/backends/neon/test/NeonEndToEndTests.cpp +++ b/src/backends/neon/test/NeonEndToEndTests.cpp @@ -6,7 +6,7 @@ #include <backendsCommon/test/EndToEndTestImpl.hpp> #include <backendsCommon/test/AbsEndToEndTestImpl.hpp> -#include <backendsCommon/test/ArithmeticTestImpl.hpp> +#include <backendsCommon/test/ComparisonEndToEndTestImpl.hpp> #include <backendsCommon/test/ConcatEndToEndTestImpl.hpp> #include <backendsCommon/test/DepthToSpaceEndToEndTestImpl.hpp> #include <backendsCommon/test/DequantizeEndToEndTestImpl.hpp> @@ -80,9 +80,9 @@ BOOST_AUTO_TEST_CASE(NeonGreaterSimpleEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(NeonGreaterSimpleEndToEndUint8Test) @@ -90,9 +90,9 @@ BOOST_AUTO_TEST_CASE(NeonGreaterSimpleEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(NeonGreaterBroadcastEndToEndTest) @@ -100,9 +100,9 @@ BOOST_AUTO_TEST_CASE(NeonGreaterBroadcastEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(NeonGreaterBroadcastEndToEndUint8Test) @@ -110,9 +110,9 @@ BOOST_AUTO_TEST_CASE(NeonGreaterBroadcastEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(NeonConcatEndToEndDim0Test) diff --git a/src/backends/reference/RefLayerSupport.cpp b/src/backends/reference/RefLayerSupport.cpp index 9342b29f47..c65886ba4d 100644 --- a/src/backends/reference/RefLayerSupport.cpp +++ b/src/backends/reference/RefLayerSupport.cpp @@ -308,6 +308,35 @@ bool RefLayerSupport::IsBatchToSpaceNdSupported(const TensorInfo& input, return supported; } +bool RefLayerSupport::IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported) const +{ + boost::ignore_unused(descriptor); + + std::array<DataType, 4> supportedInputTypes = + { + DataType::Float32, + DataType::Float16, + DataType::QuantisedAsymm8, + DataType::QuantisedSymm16 + }; + + bool supported = true; + supported &= CheckSupportRule(TypeAnyOf(input0, supportedInputTypes), reasonIfUnsupported, + "Reference comparison: input 0 is not a supported type"); + + supported &= CheckSupportRule(TypesAreEqual(input0, input1), reasonIfUnsupported, + "Reference comparison: input 0 and Input 1 types are mismatched"); + + supported &= CheckSupportRule(TypeIs(output, DataType::Boolean), reasonIfUnsupported, + "Reference comparison: output is not of type Boolean"); + + return supported; +} + bool RefLayerSupport::IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -644,29 +673,11 @@ bool RefLayerSupport::IsEqualSupported(const TensorInfo& input0, const TensorInfo& output, Optional<std::string&> reasonIfUnsupported) const { - bool supported = true; - - std::array<DataType,4> supportedTypes = - { - DataType::Float32, - DataType::Float16, - DataType::QuantisedAsymm8, - DataType::QuantisedSymm16 - }; - - supported &= CheckSupportRule(TypeAnyOf(input0, supportedTypes), reasonIfUnsupported, - "Reference equal: input 0 is not a supported type."); - - supported &= CheckSupportRule(TypeAnyOf(input1, supportedTypes), reasonIfUnsupported, - "Reference equal: input 1 is not a supported type."); - - supported &= CheckSupportRule(TypesAreEqual(input0, input1), reasonIfUnsupported, - "Reference equal: input 0 and Input 1 types are mismatched"); - - supported &= CheckSupportRule(ShapesAreBroadcastCompatible(input0, input1, output), reasonIfUnsupported, - "Reference equal: shapes are not suitable for implicit broadcast."); - - return supported; + return IsComparisonSupported(input0, + input1, + output, + ComparisonDescriptor(ComparisonOperation::Equal), + reasonIfUnsupported); } bool RefLayerSupport::IsFakeQuantizationSupported(const TensorInfo& input, @@ -802,29 +813,11 @@ bool RefLayerSupport::IsGreaterSupported(const TensorInfo& input0, const TensorInfo& output, Optional<std::string&> reasonIfUnsupported) const { - bool supported = true; - - std::array<DataType,4> supportedTypes = - { - DataType::Float32, - DataType::Float16, - DataType::QuantisedAsymm8, - DataType::QuantisedSymm16 - }; - - supported &= CheckSupportRule(TypeAnyOf(input0, supportedTypes), reasonIfUnsupported, - "Reference greater: input 0 is not a supported type."); - - supported &= CheckSupportRule(TypeAnyOf(input1, supportedTypes), reasonIfUnsupported, - "Reference greater: input 1 is not a supported type."); - - supported &= CheckSupportRule(TypesAreEqual(input0, input1), reasonIfUnsupported, - "Reference greater: input 0 and Input 1 types are mismatched"); - - supported &= CheckSupportRule(ShapesAreBroadcastCompatible(input0, input1, output), reasonIfUnsupported, - "Reference greater: shapes are not suitable for implicit broadcast."); - - return supported; + return IsComparisonSupported(input0, + input1, + output, + ComparisonDescriptor(ComparisonOperation::Greater), + reasonIfUnsupported); } bool RefLayerSupport::IsInputSupported(const TensorInfo& input, diff --git a/src/backends/reference/RefLayerSupport.hpp b/src/backends/reference/RefLayerSupport.hpp index 5c71e8d337..04b355ee0a 100644 --- a/src/backends/reference/RefLayerSupport.hpp +++ b/src/backends/reference/RefLayerSupport.hpp @@ -45,6 +45,12 @@ public: const BatchToSpaceNdDescriptor& descriptor, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsComparisonSupported(const TensorInfo& input0, + const TensorInfo& input1, + const TensorInfo& output, + const ComparisonDescriptor& descriptor, + Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + bool IsConcatSupported(const std::vector<const TensorInfo*> inputs, const TensorInfo& output, const ConcatDescriptor& descriptor, @@ -106,6 +112,7 @@ public: const TensorInfo& output, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsEqualSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& output, @@ -131,6 +138,7 @@ public: const TensorInfo& output, Optional<std::string&> reasonIfUnsupported = EmptyOptional()) const override; + ARMNN_DEPRECATED_MSG("Use IsComparisonSupported instead") bool IsGreaterSupported(const TensorInfo& input0, const TensorInfo& input1, const TensorInfo& output, diff --git a/src/backends/reference/RefWorkloadFactory.cpp b/src/backends/reference/RefWorkloadFactory.cpp index 1f6d1d7e8b..c2cb51abf3 100644 --- a/src/backends/reference/RefWorkloadFactory.cpp +++ b/src/backends/reference/RefWorkloadFactory.cpp @@ -131,6 +131,12 @@ std::unique_ptr<IWorkload> RefWorkloadFactory::CreateBatchToSpaceNd(const BatchT return std::make_unique<RefBatchToSpaceNdWorkload>(descriptor, info); } +std::unique_ptr<IWorkload> RefWorkloadFactory::CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const +{ + return std::make_unique<RefComparisonWorkload>(descriptor, info); +} + std::unique_ptr<IWorkload> RefWorkloadFactory::CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const { @@ -208,7 +214,10 @@ std::unique_ptr<IWorkload> RefWorkloadFactory::CreateDivision(const DivisionQueu std::unique_ptr<IWorkload> RefWorkloadFactory::CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return std::make_unique<RefEqualWorkload>(descriptor, info); + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters.m_Operation = ComparisonOperation::Equal; + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> RefWorkloadFactory::CreateFakeQuantization( @@ -240,7 +249,10 @@ std::unique_ptr<IWorkload> RefWorkloadFactory::CreateGather(const GatherQueueDes std::unique_ptr<IWorkload> RefWorkloadFactory::CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const { - return std::make_unique<RefGreaterWorkload>(descriptor, info); + ComparisonQueueDescriptor comparisonDescriptor; + comparisonDescriptor.m_Parameters.m_Operation = ComparisonOperation::Greater; + + return CreateComparison(comparisonDescriptor, info); } std::unique_ptr<IWorkload> RefWorkloadFactory::CreateInput(const InputQueueDescriptor& descriptor, diff --git a/src/backends/reference/RefWorkloadFactory.hpp b/src/backends/reference/RefWorkloadFactory.hpp index 41e9b28ea2..7b73d5b21f 100644 --- a/src/backends/reference/RefWorkloadFactory.hpp +++ b/src/backends/reference/RefWorkloadFactory.hpp @@ -78,6 +78,9 @@ public: std::unique_ptr<IWorkload> CreateBatchToSpaceNd(const BatchToSpaceNdQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + std::unique_ptr<IWorkload> CreateComparison(const ComparisonQueueDescriptor& descriptor, + const WorkloadInfo& info) const override; + std::unique_ptr<IWorkload> CreateConcat(const ConcatQueueDescriptor& descriptor, const WorkloadInfo& info) const override; @@ -111,6 +114,7 @@ public: std::unique_ptr<IWorkload> CreateDivision(const DivisionQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") std::unique_ptr<IWorkload> CreateEqual(const EqualQueueDescriptor& descriptor, const WorkloadInfo& info) const override; @@ -126,6 +130,7 @@ public: std::unique_ptr<IWorkload> CreateGather(const GatherQueueDescriptor& descriptor, const WorkloadInfo& info) const override; + ARMNN_DEPRECATED_MSG("Use CreateComparison instead") std::unique_ptr<IWorkload> CreateGreater(const GreaterQueueDescriptor& descriptor, const WorkloadInfo& info) const override; diff --git a/src/backends/reference/backend.mk b/src/backends/reference/backend.mk index 49b07a41d2..7e97acdee2 100644 --- a/src/backends/reference/backend.mk +++ b/src/backends/reference/backend.mk @@ -47,6 +47,7 @@ BACKEND_SOURCES := \ workloads/RefArgMinMaxWorkload.cpp \ workloads/RefBatchNormalizationWorkload.cpp \ workloads/RefBatchToSpaceNdWorkload.cpp \ + workloads/RefComparisonWorkload.cpp \ workloads/RefConcatWorkload.cpp \ workloads/RefConstantWorkload.cpp \ workloads/RefConvertFp16ToFp32Workload.cpp \ diff --git a/src/backends/reference/test/RefEndToEndTests.cpp b/src/backends/reference/test/RefEndToEndTests.cpp index 370ef6599b..1968e4da7e 100644 --- a/src/backends/reference/test/RefEndToEndTests.cpp +++ b/src/backends/reference/test/RefEndToEndTests.cpp @@ -6,8 +6,8 @@ #include <backendsCommon/test/EndToEndTestImpl.hpp> #include <backendsCommon/test/AbsEndToEndTestImpl.hpp> -#include <backendsCommon/test/ArithmeticTestImpl.hpp> #include <backendsCommon/test/BatchToSpaceNdEndToEndTestImpl.hpp> +#include <backendsCommon/test/ComparisonEndToEndTestImpl.hpp> #include <backendsCommon/test/ConcatEndToEndTestImpl.hpp> #include <backendsCommon/test/DepthToSpaceEndToEndTestImpl.hpp> #include <backendsCommon/test/DequantizeEndToEndTestImpl.hpp> @@ -348,9 +348,9 @@ BOOST_AUTO_TEST_CASE(RefEqualSimpleEndToEndTest) const std::vector<uint8_t> expectedOutput({ 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1 }); - ArithmeticSimpleEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Equal, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Equal, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefGreaterSimpleEndToEndTest) @@ -358,9 +358,9 @@ BOOST_AUTO_TEST_CASE(RefGreaterSimpleEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefEqualSimpleEndToEndUint8Test) @@ -368,9 +368,9 @@ BOOST_AUTO_TEST_CASE(RefEqualSimpleEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1 }); - ArithmeticSimpleEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Equal, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Equal, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefGreaterSimpleEndToEndUint8Test) @@ -378,9 +378,9 @@ BOOST_AUTO_TEST_CASE(RefGreaterSimpleEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticSimpleEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonSimpleEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefEqualBroadcastEndToEndTest) @@ -388,9 +388,9 @@ BOOST_AUTO_TEST_CASE(RefEqualBroadcastEndToEndTest) const std::vector<uint8_t> expectedOutput({ 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Equal, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Equal, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefGreaterBroadcastEndToEndTest) @@ -398,9 +398,9 @@ BOOST_AUTO_TEST_CASE(RefGreaterBroadcastEndToEndTest) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::Float32, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::Float32>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefEqualBroadcastEndToEndUint8Test) @@ -408,9 +408,9 @@ BOOST_AUTO_TEST_CASE(RefEqualBroadcastEndToEndUint8Test) const std::vector<uint8_t > expectedOutput({ 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Equal, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Equal, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefGreaterBroadcastEndToEndUint8Test) @@ -418,9 +418,9 @@ BOOST_AUTO_TEST_CASE(RefGreaterBroadcastEndToEndUint8Test) const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 }); - ArithmeticBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8, armnn::DataType::Boolean>(defaultBackends, - LayerType::Greater, - expectedOutput); + ComparisonBroadcastEndToEnd<armnn::DataType::QuantisedAsymm8>(defaultBackends, + ComparisonOperation::Greater, + expectedOutput); } BOOST_AUTO_TEST_CASE(RefBatchToSpaceNdEndToEndFloat32NHWCTest) diff --git a/src/backends/reference/workloads/CMakeLists.txt b/src/backends/reference/workloads/CMakeLists.txt index b8eb95c729..7844518620 100644 --- a/src/backends/reference/workloads/CMakeLists.txt +++ b/src/backends/reference/workloads/CMakeLists.txt @@ -63,6 +63,8 @@ list(APPEND armnnRefBackendWorkloads_sources RefBatchNormalizationWorkload.hpp RefBatchToSpaceNdWorkload.cpp RefBatchToSpaceNdWorkload.hpp + RefComparisonWorkload.cpp + RefComparisonWorkload.hpp RefConcatWorkload.cpp RefConcatWorkload.hpp RefConstantWorkload.cpp diff --git a/src/backends/reference/workloads/ElementwiseFunction.cpp b/src/backends/reference/workloads/ElementwiseFunction.cpp index 7a5c071f70..888037f9a6 100644 --- a/src/backends/reference/workloads/ElementwiseFunction.cpp +++ b/src/backends/reference/workloads/ElementwiseFunction.cpp @@ -32,6 +32,11 @@ template struct armnn::ElementwiseFunction<std::multiplies<float>>; template struct armnn::ElementwiseFunction<std::divides<float>>; template struct armnn::ElementwiseFunction<armnn::maximum<float>>; template struct armnn::ElementwiseFunction<armnn::minimum<float>>; + +// Comparison template struct armnn::ElementwiseFunction<std::equal_to<float>>; template struct armnn::ElementwiseFunction<std::greater<float>>; - +template struct armnn::ElementwiseFunction<std::greater_equal<float>>; +template struct armnn::ElementwiseFunction<std::less<float>>; +template struct armnn::ElementwiseFunction<std::less_equal<float>>; +template struct armnn::ElementwiseFunction<std::not_equal_to<float>>; diff --git a/src/backends/reference/workloads/RefComparisonWorkload.cpp b/src/backends/reference/workloads/RefComparisonWorkload.cpp new file mode 100644 index 0000000000..60446226be --- /dev/null +++ b/src/backends/reference/workloads/RefComparisonWorkload.cpp @@ -0,0 +1,102 @@ +// +// Copyright © 2019 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// + +#include "RefComparisonWorkload.hpp" + +#include "Decoders.hpp" +#include "ElementwiseFunction.hpp" +#include "Encoders.hpp" +#include "RefWorkloadUtils.hpp" + +#include <Profiling.hpp> + +#include <armnn/TypesUtils.hpp> + +#include <functional> + +namespace armnn +{ + +RefComparisonWorkload::RefComparisonWorkload(const ComparisonQueueDescriptor& desc, + const WorkloadInfo& info) + : BaseWorkload<ComparisonQueueDescriptor>(desc, info) +{} + +void RefComparisonWorkload::PostAllocationConfigure() +{ + const TensorInfo& inputInfo0 = GetTensorInfo(m_Data.m_Inputs[0]); + const TensorInfo& inputInfo1 = GetTensorInfo(m_Data.m_Inputs[1]); + const TensorInfo& outputInfo = GetTensorInfo(m_Data.m_Outputs[0]); + + m_Input0 = MakeDecoder<InType>(inputInfo0); + m_Input1 = MakeDecoder<InType>(inputInfo1); + + m_Output = MakeEncoder<OutType>(outputInfo); +} + +void RefComparisonWorkload::Execute() const +{ + ARMNN_SCOPED_PROFILING_EVENT(Compute::CpuRef, "RefComparisonWorkload_Execute"); + + const TensorInfo& inputInfo0 = GetTensorInfo(m_Data.m_Inputs[0]); + const TensorInfo& inputInfo1 = GetTensorInfo(m_Data.m_Inputs[1]); + const TensorInfo& outputInfo = GetTensorInfo(m_Data.m_Outputs[0]); + + const TensorShape& inShape0 = inputInfo0.GetShape(); + const TensorShape& inShape1 = inputInfo1.GetShape(); + const TensorShape& outShape = outputInfo.GetShape(); + + m_Input0->Reset(m_Data.m_Inputs[0]->Map()); + m_Input1->Reset(m_Data.m_Inputs[1]->Map()); + m_Output->Reset(m_Data.m_Outputs[0]->Map()); + + using EqualFunction = ElementwiseFunction<std::equal_to<InType>>; + using GreaterFunction = ElementwiseFunction<std::greater<InType>>; + using GreaterOrEqualFunction = ElementwiseFunction<std::greater_equal<InType>>; + using LessFunction = ElementwiseFunction<std::less<InType>>; + using LessOrEqualFunction = ElementwiseFunction<std::less_equal<InType>>; + using NotEqualFunction = ElementwiseFunction<std::not_equal_to<InType>>; + + switch (m_Data.m_Parameters.m_Operation) + { + case ComparisonOperation::Equal: + { + EqualFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + case ComparisonOperation::Greater: + { + GreaterFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + case ComparisonOperation::GreaterOrEqual: + { + GreaterOrEqualFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + case ComparisonOperation::Less: + { + LessFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + case ComparisonOperation::LessOrEqual: + { + LessOrEqualFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + case ComparisonOperation::NotEqual: + { + NotEqualFunction(inShape0, inShape1, outShape, *m_Input0, *m_Input1, *m_Output); + break; + } + default: + { + throw InvalidArgumentException(std::string("Unsupported comparison operation ") + + GetComparisonOperationAsCString(m_Data.m_Parameters.m_Operation), CHECK_LOCATION()); + } + } +} + +} // namespace armnn diff --git a/src/backends/reference/workloads/RefComparisonWorkload.hpp b/src/backends/reference/workloads/RefComparisonWorkload.hpp new file mode 100644 index 0000000000..a19e4a0540 --- /dev/null +++ b/src/backends/reference/workloads/RefComparisonWorkload.hpp @@ -0,0 +1,34 @@ +// +// Copyright © 2019 Arm Ltd. All rights reserved. +// SPDX-License-Identifier: MIT +// + +#pragma once + +#include "BaseIterator.hpp" + +#include <backendsCommon/Workload.hpp> +#include <backendsCommon/WorkloadData.hpp> + +namespace armnn +{ + +class RefComparisonWorkload : public BaseWorkload<ComparisonQueueDescriptor> +{ +public: + using BaseWorkload<ComparisonQueueDescriptor>::m_Data; + + RefComparisonWorkload(const ComparisonQueueDescriptor& descriptor, const WorkloadInfo& info); + void PostAllocationConfigure() override; + void Execute() const override; + +private: + using InType = float; + using OutType = bool; + + std::unique_ptr<Decoder<InType>> m_Input0; + std::unique_ptr<Decoder<InType>> m_Input1; + std::unique_ptr<Encoder<OutType>> m_Output; +}; + +} // namespace armnn diff --git a/src/backends/reference/workloads/RefElementwiseWorkload.cpp b/src/backends/reference/workloads/RefElementwiseWorkload.cpp index 6431348bc2..7e02f032ef 100644 --- a/src/backends/reference/workloads/RefElementwiseWorkload.cpp +++ b/src/backends/reference/workloads/RefElementwiseWorkload.cpp @@ -86,11 +86,3 @@ template class armnn::RefElementwiseWorkload<armnn::maximum<float>, template class armnn::RefElementwiseWorkload<armnn::minimum<float>, armnn::MinimumQueueDescriptor, armnn::StringMapping::RefMinimumWorkload_Execute>; - -template class armnn::RefElementwiseWorkload<std::equal_to<float>, - armnn::EqualQueueDescriptor, - armnn::StringMapping::RefEqualWorkload_Execute>; - -template class armnn::RefElementwiseWorkload<std::greater<float>, - armnn::GreaterQueueDescriptor, - armnn::StringMapping::RefGreaterWorkload_Execute>; diff --git a/src/backends/reference/workloads/RefElementwiseWorkload.hpp b/src/backends/reference/workloads/RefElementwiseWorkload.hpp index 651942e9e5..ee0d80b172 100644 --- a/src/backends/reference/workloads/RefElementwiseWorkload.hpp +++ b/src/backends/reference/workloads/RefElementwiseWorkload.hpp @@ -65,13 +65,4 @@ using RefMinimumWorkload = MinimumQueueDescriptor, StringMapping::RefMinimumWorkload_Execute>; -using RefEqualWorkload = - RefElementwiseWorkload<std::equal_to<float>, - armnn::EqualQueueDescriptor, - armnn::StringMapping::RefEqualWorkload_Execute>; - -using RefGreaterWorkload = - RefElementwiseWorkload<std::greater<float>, - armnn::GreaterQueueDescriptor, - armnn::StringMapping::RefGreaterWorkload_Execute>; } // armnn diff --git a/src/backends/reference/workloads/RefWorkloads.hpp b/src/backends/reference/workloads/RefWorkloads.hpp index 79d1935823..1f9ad4a19a 100644 --- a/src/backends/reference/workloads/RefWorkloads.hpp +++ b/src/backends/reference/workloads/RefWorkloads.hpp @@ -20,6 +20,7 @@ #include "RefArgMinMaxWorkload.hpp" #include "RefBatchNormalizationWorkload.hpp" #include "RefBatchToSpaceNdWorkload.hpp" +#include "RefComparisonWorkload.hpp" #include "RefConvolution2dWorkload.hpp" #include "RefConstantWorkload.hpp" #include "RefConcatWorkload.hpp" diff --git a/src/backends/reference/workloads/StringMapping.hpp b/src/backends/reference/workloads/StringMapping.hpp index 073a5a6833..1654b78088 100644 --- a/src/backends/reference/workloads/StringMapping.hpp +++ b/src/backends/reference/workloads/StringMapping.hpp @@ -18,9 +18,7 @@ struct StringMapping public: enum Id { RefAdditionWorkload_Execute, - RefEqualWorkload_Execute, RefDivisionWorkload_Execute, - RefGreaterWorkload_Execute, RefMaximumWorkload_Execute, RefMinimumWorkload_Execute, RefMultiplicationWorkload_Execute, @@ -40,8 +38,6 @@ private: { m_Strings[RefAdditionWorkload_Execute] = "RefAdditionWorkload_Execute"; m_Strings[RefDivisionWorkload_Execute] = "RefDivisionWorkload_Execute"; - m_Strings[RefEqualWorkload_Execute] = "RefEqualWorkload_Execute"; - m_Strings[RefGreaterWorkload_Execute] = "RefGreaterWorkload_Execute"; m_Strings[RefMaximumWorkload_Execute] = "RefMaximumWorkload_Execute"; m_Strings[RefMinimumWorkload_Execute] = "RefMinimumWorkload_Execute"; m_Strings[RefMultiplicationWorkload_Execute] = "RefMultiplicationWorkload_Execute"; |