src/backends/neon/test/NeonEndToEndTests.cpp

//
// Copyright © 2017 Arm Ltd. All rights reserved.
// SPDX-License-Identifier: MIT
//

#include <backendsCommon/test/EndToEndTestImpl.hpp>

#include <backendsCommon/test/ActivationEndToEndTestImpl.hpp>
#include <backendsCommon/test/ArgMinMaxEndToEndTestImpl.hpp>
#include <backendsCommon/test/ComparisonEndToEndTestImpl.hpp>
#include <backendsCommon/test/ConcatEndToEndTestImpl.hpp>
#include <backendsCommon/test/DepthToSpaceEndToEndTestImpl.hpp>
#include <backendsCommon/test/DequantizeEndToEndTestImpl.hpp>
#include <backendsCommon/test/DetectionPostProcessEndToEndTestImpl.hpp>
#include <backendsCommon/test/ElementwiseUnaryEndToEndTestImpl.hpp>
#include <backendsCommon/test/FillEndToEndTestImpl.hpp>
#include <backendsCommon/test/InstanceNormalizationEndToEndTestImpl.hpp>
#include <backendsCommon/test/PreluEndToEndTestImpl.hpp>
#include <backendsCommon/test/QLstmEndToEndTestImpl.hpp>
#include <backendsCommon/test/QuantizedLstmEndToEndTestImpl.hpp>
#include <backendsCommon/test/SpaceToDepthEndToEndTestImpl.hpp>
#include <backendsCommon/test/SplitterEndToEndTestImpl.hpp>
#include <backendsCommon/test/TransposeConvolution2dEndToEndTestImpl.hpp>

#include <doctest/doctest.h>

TEST_SUITE("NeonEndToEnd")
{
std::vector<armnn::BackendId> neonDefaultBackends = {armnn::Compute::CpuAcc};

// Abs
TEST_CASE("NeonAbsEndToEndTestFloat32")
{
    std::vector<float> expectedOutput =
    {
        1.f, 1.f, 1.f, 1.f, 5.f, 5.f, 5.f, 5.f,
        3.f, 3.f, 3.f, 3.f, 4.f, 4.f, 4.f, 4.f
    };

    ElementwiseUnarySimpleEndToEnd<armnn::DataType::Float32>(neonDefaultBackends,
                                                             UnaryOperation::Abs,
                                                             expectedOutput);
}

// Constant
TEST_CASE("ConstantUsage_Neon_Float32")
{
    CHECK(ConstantUsageFloat32Test(neonDefaultBackends));
}

#if defined(ARMNNREF_ENABLED)

// This test unit needs the reference backend, it's not available if the reference backend is not built

TEST_CASE("FallbackToCpuRef")
{
    using namespace armnn;

    // Create runtime in which test will run and allow fallback to CpuRef.
    IRuntime::CreationOptions options;
    IRuntimePtr runtime(IRuntime::Create(options));

    // Builds up the structure of the network.
    INetworkPtr net(INetwork::Create());

    IConnectableLayer* input = net->AddInputLayer(0);

    // This layer configuration isn't supported by CpuAcc but we allow fallback to CpuRef so it shoud pass.
    NormalizationDescriptor descriptor;
    IConnectableLayer* pooling = net->AddNormalizationLayer(descriptor);

    IConnectableLayer* output = net->AddOutputLayer(0);

    input->GetOutputSlot(0).Connect(pooling->GetInputSlot(0));
    pooling->GetOutputSlot(0).Connect(output->GetInputSlot(0));

    input->GetOutputSlot(0).SetTensorInfo(TensorInfo({ 1, 1, 4, 4 }, DataType::Float32));
    pooling->GetOutputSlot(0).SetTensorInfo(TensorInfo({ 1, 1, 4, 4 }, DataType::Float32));

    // optimize the network
    std::vector<BackendId> backends = {Compute::CpuAcc, Compute::CpuRef};
    IOptimizedNetworkPtr optNet = Optimize(*net, backends, runtime->GetDeviceSpec());

    // Load it into the runtime. It should pass.
    NetworkId netId;
    CHECK(runtime->LoadNetwork(netId, std::move(optNet)) == Status::Success);
}

#endif

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 });

    ComparisonSimpleEndToEnd<armnn::DataType::Float32>(neonDefaultBackends,
                                                       ComparisonOperation::Greater,
                                                       expectedOutput);
}

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 });

    ComparisonSimpleEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends,
                                                               ComparisonOperation::Greater,
                                                               expectedOutput);
}

TEST_CASE("NeonGreaterBroadcastEndToEndTest")
{
    const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1,
                                                1, 1, 1, 1, 1, 1 });

    ComparisonBroadcastEndToEnd<armnn::DataType::Float32>(neonDefaultBackends,
                                                          ComparisonOperation::Greater,
                                                          expectedOutput);
}

TEST_CASE("NeonGreaterBroadcastEndToEndUint8Test")
{
    const std::vector<uint8_t> expectedOutput({ 0, 1, 0, 0, 0, 1,
                                                1, 1, 1, 1, 1, 1 });

    ComparisonBroadcastEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends,
                                                                  ComparisonOperation::Greater,
                                                                  expectedOutput);
}

TEST_CASE("NeonConcatEndToEndDim0Test")
{
    ConcatDim0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonConcatEndToEndDim0Uint8Test")
{
    ConcatDim0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonConcatEndToEndDim1Test")
{
    ConcatDim1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonConcatEndToEndDim1Uint8Test")
{
    ConcatDim1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonConcatEndToEndDim3Test")
{
    ConcatDim3EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonConcatEndToEndDim3Uint8Test")
{
    ConcatDim3EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

// DepthToSpace
TEST_CASE("DephtToSpaceEndToEndNchwFloat32")
{
    DepthToSpaceEndToEnd<armnn::DataType::Float32>(neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("DephtToSpaceEndToEndNchwFloat16")
{
    DepthToSpaceEndToEnd<armnn::DataType::Float16>(neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("DephtToSpaceEndToEndNchwUint8")
{
    DepthToSpaceEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("DephtToSpaceEndToEndNchwInt16")
{
    DepthToSpaceEndToEnd<armnn::DataType::QSymmS16>(neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("DephtToSpaceEndToEndNhwcFloat32")
{
    DepthToSpaceEndToEnd<armnn::DataType::Float32>(neonDefaultBackends, armnn::DataLayout::NHWC);
}

TEST_CASE("DephtToSpaceEndToEndNhwcFloat16")
{
    DepthToSpaceEndToEnd<armnn::DataType::Float16>(neonDefaultBackends, armnn::DataLayout::NHWC);
}

TEST_CASE("DephtToSpaceEndToEndNhwcUint8")
{
    DepthToSpaceEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends, armnn::DataLayout::NHWC);
}

TEST_CASE("DephtToSpaceEndToEndNhwcInt16")
{
    DepthToSpaceEndToEnd<armnn::DataType::QSymmS16>(neonDefaultBackends, armnn::DataLayout::NHWC);
}

// Dequantize
TEST_CASE("DequantizeEndToEndSimpleTest")
{
    DequantizeEndToEndSimple<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("DequantizeEndToEndOffsetTest")
{
    DequantizeEndToEndOffset<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonEluEndToEndTestFloat32")
{
    EluEndToEndTest<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonEluEndToEndTestFloat16")
{
    EluEndToEndTest<armnn::DataType::Float16>(neonDefaultBackends);
}

// HardSwish
TEST_CASE("NeonHardSwishEndToEndTestFloat32")
{
    HardSwishEndToEndTest<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonHardSwishEndToEndTestFloat16")
{
    HardSwishEndToEndTest<armnn::DataType::Float16>(neonDefaultBackends);
}

TEST_CASE("NeonHardSwishEndToEndTestQAsymmS8")
{
    HardSwishEndToEndTest<armnn::DataType::QAsymmS8>(neonDefaultBackends);
}

TEST_CASE("NeonHardSwishEndToEndTestQAsymmU8")
{
    HardSwishEndToEndTest<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonPreluEndToEndFloat32Test")
{
    PreluEndToEndNegativeTest<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonPreluEndToEndTestUint8Test")
{
    PreluEndToEndPositiveTest<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSpaceToDepthNhwcEndToEndTest1")
{
    SpaceToDepthNhwcEndToEndTest1(neonDefaultBackends);
}

TEST_CASE("NeonSpaceToDepthNchwEndToEndTest1")
{
    SpaceToDepthNchwEndToEndTest1(neonDefaultBackends);
}

TEST_CASE("NeonSpaceToDepthNhwcEndToEndTest2")
{
    SpaceToDepthNhwcEndToEndTest2(neonDefaultBackends);
}

TEST_CASE("NeonSpaceToDepthNchwEndToEndTest2")
{
    SpaceToDepthNchwEndToEndTest2(neonDefaultBackends);
}

TEST_CASE("NeonSplitter1dEndToEndTest")
{
    Splitter1dEndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter1dEndToEndUint8Test")
{
    Splitter1dEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter2dDim0EndToEndTest")
{
    Splitter2dDim0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter2dDim1EndToEndTest")
{
    Splitter2dDim1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter2dDim0EndToEndUint8Test")
{
    Splitter2dDim0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter2dDim1EndToEndUint8Test")
{
    Splitter2dDim1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim0EndToEndTest")
{
    Splitter3dDim0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim1EndToEndTest")
{
    Splitter3dDim1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim2EndToEndTest")
{
    Splitter3dDim2EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim0EndToEndUint8Test")
{
    Splitter3dDim0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim1EndToEndUint8Test")
{
    Splitter3dDim1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter3dDim2EndToEndUint8Test")
{
    Splitter3dDim2EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim0EndToEndTest")
{
    Splitter4dDim0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim1EndToEndTest")
{
    Splitter4dDim1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim2EndToEndTest")
{
    Splitter4dDim2EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim3EndToEndTest")
{
    Splitter4dDim3EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim0EndToEndUint8Test")
{
    Splitter4dDim0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim1EndToEndUint8Test")
{
    Splitter4dDim1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim2EndToEndUint8Test")
{
    Splitter4dDim2EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonSplitter4dDim3EndToEndUint8Test")
{
    Splitter4dDim3EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonQuantizedLstmEndToEndTest")
{
    QuantizedLstmEndToEnd(neonDefaultBackends);
}

TEST_CASE("NeonTransposeConvolution2dEndToEndFloatNchwTest")
{
    TransposeConvolution2dEndToEnd<armnn::DataType::Float32, armnn::DataType::Float32>(
        neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("NeonTransposeConvolution2dEndToEndUint8NchwTest")
{
    TransposeConvolution2dEndToEnd<armnn::DataType::QAsymmU8, armnn::DataType::Signed32>(
        neonDefaultBackends, armnn::DataLayout::NCHW);
}

TEST_CASE("NeonTransposeConvolution2dEndToEndFloatNhwcTest")
{
    TransposeConvolution2dEndToEnd<armnn::DataType::Float32, armnn::DataType::Float32>(
        neonDefaultBackends, armnn::DataLayout::NHWC);
}

TEST_CASE("NeonTransposeConvolution2dEndToEndUint8NhwcTest")
{
    TransposeConvolution2dEndToEnd<armnn::DataType::QAsymmU8, armnn::DataType::Signed32>(
        neonDefaultBackends, armnn::DataLayout::NHWC);
}

TEST_CASE("NeonImportNonAlignedInputPointerTest")
{
    ImportNonAlignedInputPointerTest(neonDefaultBackends);
}

TEST_CASE("NeonExportNonAlignedOutputPointerTest")
{
    ExportNonAlignedOutputPointerTest(neonDefaultBackends);
}

TEST_CASE("NeonImportAlignedPointerTest")
{
    ImportAlignedPointerTest(neonDefaultBackends);
}

TEST_CASE("NeonImportOnlyWorkload")
{
    ImportOnlyWorkload(neonDefaultBackends);
}

TEST_CASE("NeonExportOnlyWorkload")
{
    ExportOnlyWorkload(neonDefaultBackends);
}

TEST_CASE("NeonImportAndExportWorkload")
{
    ImportAndExportWorkload(neonDefaultBackends);
}

TEST_CASE("NeonExportOutputWithSeveralOutputSlotConnectionsTest")
{
    ExportOutputWithSeveralOutputSlotConnectionsTest(neonDefaultBackends);
}

// InstanceNormalization
TEST_CASE("NeonInstanceNormalizationNchwEndToEndTest1")
{
    InstanceNormalizationNchwEndToEndTest1(neonDefaultBackends);
}

TEST_CASE("NeonInstanceNormalizationNchwEndToEndTest2")
{
    InstanceNormalizationNchwEndToEndTest2(neonDefaultBackends);
}

// Fill
TEST_CASE("NeonFillEndToEndTest")
{
    FillEndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("RefFillEndToEndTestFloat16")
{
    FillEndToEnd<armnn::DataType::Float16>(neonDefaultBackends);
}

TEST_CASE("NeonFillEndToEndTestInt32")
{
    FillEndToEnd<armnn::DataType::Signed32>(neonDefaultBackends);
}

// ArgMinMax
TEST_CASE("NeonArgMaxSimpleTest")
{
    ArgMaxEndToEndSimple<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinSimpleTest")
{
    ArgMinEndToEndSimple<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis0Test")
{
    ArgMaxAxis0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis0Test")
{
    ArgMinAxis0EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis1Test")
{
    ArgMaxAxis1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis1Test")
{
    ArgMinAxis1EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis2Test")
{
    ArgMaxAxis2EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis2Test")
{
    ArgMinAxis2EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis3Test")
{
    ArgMaxAxis3EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis3Test")
{
    ArgMinAxis3EndToEnd<armnn::DataType::Float32>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxSimpleTestQuantisedAsymm8")
{
    ArgMaxEndToEndSimple<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinSimpleTestQuantisedAsymm8")
{
    ArgMinEndToEndSimple<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis0TestQuantisedAsymm8")
{
    ArgMaxAxis0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis0TestQuantisedAsymm8")
{
    ArgMinAxis0EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis1TestQuantisedAsymm8")
{
    ArgMaxAxis1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis1TestQuantisedAsymm8")
{
    ArgMinAxis1EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis2TestQuantisedAsymm8")
{
    ArgMaxAxis2EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis2TestQuantisedAsymm8")
{
    ArgMinAxis2EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMaxAxis3TestQuantisedAsymm8")
{
    ArgMaxAxis3EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonArgMinAxis3TestQuantisedAsymm8")
{
    ArgMinAxis3EndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends);
}

TEST_CASE("NeonStridedSliceInvalidSliceEndToEndTest")
{
    StridedSliceInvalidSliceEndToEndTest(neonDefaultBackends);
}

TEST_CASE("NeonForceImportWithAlignedBuffersEndToEndTest"
          // Currently, the Neon workload for activation does not support tensor handle replacement so this test case
          // will always fail.
          * doctest::skip(true))
{
    ForceImportWithAlignedBuffersEndToEndTest(neonDefaultBackends);
}

TEST_CASE("NeonForceImportWithMisalignedInputBuffersEndToEndTest"
          // Currently, the Neon workload for activation does not support tensor handle replacement so this test case
          // will always fail.
          * doctest::skip(true))
{
    ForceImportWithMisalignedInputBuffersEndToEndTest(neonDefaultBackends);
}

TEST_CASE("NeonForceImportWithMisalignedOutputBuffersEndToEndTest"
          // Currently, the Neon workload for activation does not support tensor handle replacement so this test case
          // will always fail.
          * doctest::skip(true))
{
    ForceImportWithMisalignedOutputBuffersEndToEndTest(neonDefaultBackends);
}

TEST_CASE("NeonForceImportWithMisalignedInputAndOutputBuffersEndToEndTest")
{
    ForceImportWithMisalignedInputAndOutputBuffersEndToEndTest(neonDefaultBackends);
}

// DISABLED
//TEST_CASE("NeonDetectionPostProcessRegularNmsTest")
//{
//    std::vector<float> boxEncodings({
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, -1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f
//                                    });
//    std::vector<float> scores({
//                                  0.0f, 0.9f, 0.8f,
//                                  0.0f, 0.75f, 0.72f,
//                                  0.0f, 0.6f, 0.5f,
//                                  0.0f, 0.93f, 0.95f,
//                                  0.0f, 0.5f, 0.4f,
//                                  0.0f, 0.3f, 0.2f
//                              });
//    std::vector<float> anchors({
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 100.5f, 1.0f, 1.0f
//                               });
//    DetectionPostProcessRegularNmsEndToEnd<armnn::DataType::Float32>(neonDefaultBackends,
//                                                                     boxEncodings,
//                                                                     scores,
//                                                                     anchors);
//}

inline void QuantizeData(uint8_t* quant, const float* dequant, const TensorInfo& info)
{
    for (size_t i = 0; i < info.GetNumElements(); i++)
    {
        quant[i] = armnn::Quantize<uint8_t>(dequant[i], info.GetQuantizationScale(), info.GetQuantizationOffset());
    }
}

// DISABLED
//TEST_CASE("NeonDetectionPostProcessRegularNmsUint8Test")
//{
//    armnn::TensorInfo boxEncodingsInfo({ 1, 6, 4 }, armnn::DataType::Float32);
//    armnn::TensorInfo scoresInfo({ 1, 6, 3 }, armnn::DataType::Float32);
//    armnn::TensorInfo anchorsInfo({ 6, 4 }, armnn::DataType::Float32);
//
//    boxEncodingsInfo.SetQuantizationScale(1.0f);
//    boxEncodingsInfo.SetQuantizationOffset(1);
//    scoresInfo.SetQuantizationScale(0.01f);
//    scoresInfo.SetQuantizationOffset(0);
//    anchorsInfo.SetQuantizationScale(0.5f);
//    anchorsInfo.SetQuantizationOffset(0);
//
//    std::vector<float> boxEncodings({
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, -1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f
//                                    });
//    std::vector<float> scores({
//                                  0.0f, 0.9f, 0.8f,
//                                  0.0f, 0.75f, 0.72f,
//                                  0.0f, 0.6f, 0.5f,
//                                  0.0f, 0.93f, 0.95f,
//                                  0.0f, 0.5f, 0.4f,
//                                  0.0f, 0.3f, 0.2f
//                              });
//    std::vector<float> anchors({
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 100.5f, 1.0f, 1.0f
//                               });
//
//    std::vector<uint8_t> qBoxEncodings(boxEncodings.size(), 0);
//    std::vector<uint8_t> qScores(scores.size(), 0);
//    std::vector<uint8_t> qAnchors(anchors.size(), 0);
//    QuantizeData(qBoxEncodings.data(), boxEncodings.data(), boxEncodingsInfo);
//    QuantizeData(qScores.data(), scores.data(), scoresInfo);
//    QuantizeData(qAnchors.data(), anchors.data(), anchorsInfo);
//    DetectionPostProcessRegularNmsEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends, qBoxEncodings,
//                                                                             qScores, qAnchors,
//                                                                             1.0f, 1, 0.01f, 0, 0.5f, 0);
//}
//
//TEST_CASE("NeonDetectionPostProcessFastNmsTest")
//{
//    std::vector<float> boxEncodings({
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, -1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f
//                                    });
//    std::vector<float> scores({
//                                  0.0f, 0.9f, 0.8f,
//                                  0.0f, 0.75f, 0.72f,
//                                  0.0f, 0.6f, 0.5f,
//                                  0.0f, 0.93f, 0.95f,
//                                  0.0f, 0.5f, 0.4f,
//                                  0.0f, 0.3f, 0.2f
//                              });
//    std::vector<float> anchors({
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 100.5f, 1.0f, 1.0f
//                               });
//    DetectionPostProcessFastNmsEndToEnd<armnn::DataType::Float32>(neonDefaultBackends,
//                                                                  boxEncodings,
//                                                                  scores,
//                                                                  anchors);
//}
//
// DISABLED
//TEST_CASE("NeonDetectionPostProcessFastNmsUint8Test")
//{
//    armnn::TensorInfo boxEncodingsInfo({ 1, 6, 4 }, armnn::DataType::Float32);
//    armnn::TensorInfo scoresInfo({ 1, 6, 3 }, armnn::DataType::Float32);
//    armnn::TensorInfo anchorsInfo({ 6, 4 }, armnn::DataType::Float32);
//
//    boxEncodingsInfo.SetQuantizationScale(1.0f);
//    boxEncodingsInfo.SetQuantizationOffset(1);
//    scoresInfo.SetQuantizationScale(0.01f);
//    scoresInfo.SetQuantizationOffset(0);
//    anchorsInfo.SetQuantizationScale(0.5f);
//    anchorsInfo.SetQuantizationOffset(0);
//
//    std::vector<float> boxEncodings({
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, -1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f,
//                                        0.0f, 1.0f, 0.0f, 0.0f,
//                                        0.0f, 0.0f, 0.0f, 0.0f
//                                    });
//    std::vector<float> scores({
//                                  0.0f, 0.9f, 0.8f,
//                                  0.0f, 0.75f, 0.72f,
//                                  0.0f, 0.6f, 0.5f,
//                                  0.0f, 0.93f, 0.95f,
//                                  0.0f, 0.5f, 0.4f,
//                                  0.0f, 0.3f, 0.2f
//                              });
//    std::vector<float> anchors({
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 0.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 10.5f, 1.0f, 1.0f,
//                                   0.5f, 100.5f, 1.0f, 1.0f
//                               });
//
//    std::vector<uint8_t> qBoxEncodings(boxEncodings.size(), 0);
//    std::vector<uint8_t> qScores(scores.size(), 0);
//    std::vector<uint8_t> qAnchors(anchors.size(), 0);
//    QuantizeData(qBoxEncodings.data(), boxEncodings.data(), boxEncodingsInfo);
//    QuantizeData(qScores.data(), scores.data(), scoresInfo);
//    QuantizeData(qAnchors.data(), anchors.data(), anchorsInfo);
//    DetectionPostProcessFastNmsEndToEnd<armnn::DataType::QAsymmU8>(neonDefaultBackends, qBoxEncodings,
//                                                                          qScores, qAnchors,
//                                                                          1.0f, 1, 0.01f, 0, 0.5f, 0);
//}

TEST_CASE("NeonQLstmEndToEndTest")
{
    QLstmEndToEnd(neonDefaultBackends);
}

}