potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-30 07:43:03 +04:00
Code Packages Releases Wiki Activity

Merge remote-tracking branch 'upstream/3.4' into merge-3.4

Browse Source
This commit is contained in:
Alexander Alekhin
2019-12-02 16:18:07 +03:00
parent 78c5e41c23 7523c777c5
commit 4b0132ed7a
51 changed files with 3276 additions and 496 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/test/test_layers.cpp
+134 -41
View File
@@ -146,8 +146,11 @@ TEST_P(Test_Caffe_layers, DeConvolution)
TEST_P(Test_Caffe_layers, InnerProduct)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NGRAPH);
if (backend == DNN_BACKEND_OPENCV && target == DNN_TARGET_OPENCL_FP16)
applyTestTag(CV_TEST_TAG_DNN_SKIP_OPENCL_FP16);
testLayerUsingCaffeModels("layer_inner_product", true);
@@ -238,12 +241,20 @@ TEST_P(Test_Caffe_layers, Concat)
{
#if defined(INF_ENGINE_RELEASE)
#if INF_ENGINE_VER_MAJOR_GE(2019010000) && INF_ENGINE_VER_MAJOR_LT(2019020000)
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD, CV_TEST_TAG_DNN_SKIP_IE_2019R1, CV_TEST_TAG_DNN_SKIP_IE_2019R1_1);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD, CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER, CV_TEST_TAG_DNN_SKIP_IE_VERSION);
#elif INF_ENGINE_VER_MAJOR_EQ(2019020000)
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_OPENCL)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_OPENCL, CV_TEST_TAG_DNN_SKIP_IE_2019R2);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 &&
(target == DNN_TARGET_OPENCL || target == DNN_TARGET_OPENCL_FP16))
applyTestTag(target == DNN_TARGET_OPENCL ? CV_TEST_TAG_DNN_SKIP_IE_OPENCL : CV_TEST_TAG_DNN_SKIP_IE_OPENCL_FP16,
CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER, CV_TEST_TAG_DNN_SKIP_IE_VERSION);
#endif
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH &&
(target == DNN_TARGET_OPENCL || target == DNN_TARGET_OPENCL_FP16))
applyTestTag(target == DNN_TARGET_OPENCL ? CV_TEST_TAG_DNN_SKIP_IE_OPENCL : CV_TEST_TAG_DNN_SKIP_IE_OPENCL_FP16,
CV_TEST_TAG_DNN_SKIP_IE_NGRAPH, CV_TEST_TAG_DNN_SKIP_IE_VERSION);
#endif
testLayerUsingCaffeModels("layer_concat");
testLayerUsingCaffeModels("layer_concat_optim", true, false);
@@ -252,8 +263,9 @@ TEST_P(Test_Caffe_layers, Concat)
TEST_P(Test_Caffe_layers, Fused_Concat)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE && (target == DNN_TARGET_OPENCL || target == DNN_TARGET_OPENCL_FP16))
applyTestTag(target == DNN_TARGET_OPENCL ? CV_TEST_TAG_DNN_SKIP_IE_OPENCL : CV_TEST_TAG_DNN_SKIP_IE_OPENCL_FP16);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && (target == DNN_TARGET_OPENCL || target == DNN_TARGET_OPENCL_FP16))
applyTestTag(target == DNN_TARGET_OPENCL ? CV_TEST_TAG_DNN_SKIP_IE_OPENCL : CV_TEST_TAG_DNN_SKIP_IE_OPENCL_FP16,
CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER, CV_TEST_TAG_DNN_SKIP_IE_VERSION);
checkBackend();
@@ -297,14 +309,15 @@ TEST_P(Test_Caffe_layers, Fused_Concat)
TEST_P(Test_Caffe_layers, Eltwise)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD);
testLayerUsingCaffeModels("layer_eltwise");
}
TEST_P(Test_Caffe_layers, PReLU)
{
testLayerUsingCaffeModels("layer_prelu", true);
double lInf = (target == DNN_TARGET_MYRIAD || target == DNN_TARGET_OPENCL_FP16) ? 0.021 : 0.0;
testLayerUsingCaffeModels("layer_prelu", true, true, 0.0, lInf);
}
// TODO: fix an unstable test case
@@ -320,8 +333,10 @@ TEST_P(Test_Caffe_layers, layer_prelu_fc)
TEST_P(Test_Caffe_layers, Reshape_Split_Slice)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NGRAPH);
Net net = readNetFromCaffe(_tf("reshape_and_slice_routines.prototxt"));
ASSERT_FALSE(net.empty());
@@ -342,8 +357,8 @@ TEST_P(Test_Caffe_layers, Reshape_Split_Slice)
TEST_P(Test_Caffe_layers, Conv_Elu)
{
#if defined(INF_ENGINE_RELEASE) && INF_ENGINE_RELEASE <= 2018050000
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD, CV_TEST_TAG_DNN_SKIP_IE_2018R5);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD, CV_TEST_TAG_DNN_SKIP_IE_VERSION);
#endif
Net net = readNetFromTensorflow(_tf("layer_elu_model.pb"));
@@ -556,29 +571,38 @@ TEST_F(Layer_RNN_Test, get_set_test)
EXPECT_EQ(shape(outputs[1]), shape(nT, nS, nH));
}
TEST(Layer_Test_ROIPooling, Accuracy)
TEST_P(Test_Caffe_layers, ROIPooling_Accuracy)
{
Net net = readNetFromCaffe(_tf("net_roi_pooling.prototxt"));
ASSERT_FALSE(net.empty());
Mat inp = blobFromNPY(_tf("net_roi_pooling.input.npy"));
Mat rois = blobFromNPY(_tf("net_roi_pooling.rois.npy"));
Mat ref = blobFromNPY(_tf("net_roi_pooling.npy"));
checkBackend(&inp, &ref);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(inp, "input");
net.setInput(rois, "rois");
net.setPreferableBackend(DNN_BACKEND_OPENCV);
Mat out = net.forward();
normAssert(out, ref);
double l1 = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 1e-3 : 1e-5;
double lInf = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 1e-3 : 1e-4;
normAssert(out, ref, "", l1, lInf);
}
TEST_P(Test_Caffe_layers, FasterRCNN_Proposal)
{
if (backend == DNN_BACKEND_OPENCV && target == DNN_TARGET_OPENCL_FP16)
applyTestTag(CV_TEST_TAG_DNN_SKIP_OPENCL_FP16);
if (backend == DNN_BACKEND_INFERENCE_ENGINE)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_NGRAPH);
if(backend == DNN_BACKEND_CUDA)
applyTestTag(CV_TEST_TAG_DNN_SKIP_CUDA); /* Proposal layer is unsupported */
@@ -812,16 +836,21 @@ TEST_P(Test_Caffe_layers, PriorBox_repeated)
randu(shape, -1.0f, 1.0f);
net.setInput(inp, "data");
net.setInput(shape, "shape");
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat out = net.forward();
Mat ref = blobFromNPY(_tf("priorbox_output.npy"));
normAssert(out, ref, "");
double l1 = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 1e-3 : 1e-5;
double lInf = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 1e-3 : 1e-4;
normAssert(out, ref, "", l1, lInf);
}
// Test PriorBoxLayer in case of no aspect ratios (just squared proposals).
TEST_P(Test_Caffe_layers, PriorBox_squares)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD);
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD, CV_TEST_TAG_DNN_SKIP_IE_NN_BUILDER);
LayerParams lp;
lp.name = "testPriorBox";
lp.type = "PriorBox";
@@ -975,10 +1004,21 @@ INSTANTIATE_TEST_CASE_P(/**/, Layer_Test_DWconv_Prelu, Combine(Values(3, 6), Val
// Using Intel's Model Optimizer generate .xml and .bin files:
// ./ModelOptimizer -w /path/to/caffemodel -d /path/to/prototxt \
// -p FP32 -i -b ${batch_size} -o /path/to/output/folder
typedef testing::TestWithParam<Target> Layer_Test_Convolution_DLDT;
typedef testing::TestWithParam<tuple<Backend, Target> > Layer_Test_Convolution_DLDT;
TEST_P(Layer_Test_Convolution_DLDT, Accuracy)
{
Target targetId = GetParam();
const Backend backendId = get<0>(GetParam());
const Target targetId = get<1>(GetParam());
if (backendId != DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && backendId != DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
throw SkipTestException("No support for async forward");
if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_API);
else if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NGRAPH);
else
FAIL() << "Unknown backendId";
std::string suffix = (targetId == DNN_TARGET_OPENCL_FP16 || targetId == DNN_TARGET_MYRIAD) ? "_fp16" : "";
Net netDefault = readNet(_tf("layer_convolution.caffemodel"), _tf("layer_convolution.prototxt"));
@@ -991,6 +1031,7 @@ TEST_P(Layer_Test_Convolution_DLDT, Accuracy)
Mat outDefault = netDefault.forward();
net.setInput(inp);
net.setPreferableBackend(backendId);
net.setPreferableTarget(targetId);
Mat out = net.forward();
@@ -1006,10 +1047,22 @@ TEST_P(Layer_Test_Convolution_DLDT, Accuracy)
TEST_P(Layer_Test_Convolution_DLDT, setInput_uint8)
{
Target targetId = GetParam();
Mat inp = blobFromNPY(_tf("blob.npy"));
const Backend backendId = get<0>(GetParam());
const Target targetId = get<1>(GetParam());
if (backendId != DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && backendId != DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
throw SkipTestException("No support for async forward");
if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_API);
else if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NGRAPH);
else
FAIL() << "Unknown backendId";
int blobSize[] = {2, 6, 75, 113};
Mat inputs[] = {Mat(4, &blobSize[0], CV_8U), Mat()};
Mat inputs[] = {Mat(inp.dims, inp.size, CV_8U), Mat()};
randu(inputs[0], 0, 255);
inputs[0].convertTo(inputs[1], CV_32F);
@@ -1019,6 +1072,7 @@ TEST_P(Layer_Test_Convolution_DLDT, setInput_uint8)
for (int i = 0; i < 2; ++i)
{
Net net = readNet(_tf("layer_convolution" + suffix + ".xml"), _tf("layer_convolution" + suffix + ".bin"));
net.setPreferableBackend(backendId);
net.setPreferableTarget(targetId);
net.setInput(inputs[i]);
outs[i] = net.forward();
@@ -1030,7 +1084,19 @@ TEST_P(Layer_Test_Convolution_DLDT, setInput_uint8)
TEST_P(Layer_Test_Convolution_DLDT, multithreading)
{
Target targetId = GetParam();
const Backend backendId = get<0>(GetParam());
const Target targetId = get<1>(GetParam());
if (backendId != DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && backendId != DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
throw SkipTestException("No support for async forward");
if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_API);
else if (backendId == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH)
setInferenceEngineBackendType(CV_DNN_BACKEND_INFERENCE_ENGINE_NGRAPH);
else
FAIL() << "Unknown backendId";
std::string suffix = (targetId == DNN_TARGET_OPENCL_FP16 || targetId == DNN_TARGET_MYRIAD) ? "_fp16" : "";
std::string xmlPath = _tf("layer_convolution" + suffix + ".xml");
std::string binPath = _tf("layer_convolution" + suffix + ".bin");
@@ -1040,7 +1106,9 @@ TEST_P(Layer_Test_Convolution_DLDT, multithreading)
firstNet.setInput(inp);
secondNet.setInput(inp);
firstNet.setPreferableBackend(backendId);
firstNet.setPreferableTarget(targetId);
secondNet.setPreferableBackend(backendId);
secondNet.setPreferableTarget(targetId);
Mat out1, out2;
@@ -1058,7 +1126,8 @@ TEST_P(Layer_Test_Convolution_DLDT, multithreading)
}
INSTANTIATE_TEST_CASE_P(/**/, Layer_Test_Convolution_DLDT,
testing::ValuesIn(getAvailableTargets(DNN_BACKEND_INFERENCE_ENGINE)));
dnnBackendsAndTargetsIE()
);
// 1. Create a .prototxt file with the following network:
// layer {
@@ -1117,17 +1186,18 @@ std::vector< std::vector<int> > list_sizes{ {1, 2, 3}, {3, 2, 1}, {5, 5, 5}, {13
INSTANTIATE_TEST_CASE_P(/*nothing*/, Test_DLDT_two_inputs_3dim, Combine(
Values(CV_8U, CV_32F), Values(CV_8U, CV_32F),
testing::ValuesIn(getAvailableTargets(DNN_BACKEND_INFERENCE_ENGINE)),
testing::ValuesIn(getAvailableTargets(DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019)),
testing::ValuesIn(list_sizes)
));
typedef testing::TestWithParam<tuple<int, int, Target> > Test_DLDT_two_inputs;
typedef testing::TestWithParam<tuple<int, int, tuple<Backend, Target> > > Test_DLDT_two_inputs;
TEST_P(Test_DLDT_two_inputs, as_backend)
{
static const float kScale = 0.5f;
static const float kScaleInv = 1.0f / kScale;
Target targetId = get<2>(GetParam());
Backend backendId = get<0>(get<2>(GetParam()));
Target targetId = get<1>(get<2>(GetParam()));
Net net;
LayerParams lp;
@@ -1146,7 +1216,7 @@ TEST_P(Test_DLDT_two_inputs, as_backend)
net.setInputsNames({"data", "second_input"});
net.setInput(firstInp, "data", kScale);
net.setInput(secondInp, "second_input", kScaleInv);
net.setPreferableBackend(DNN_BACKEND_INFERENCE_ENGINE);
net.setPreferableBackend(backendId);
net.setPreferableTarget(targetId);
Mat out = net.forward();
@@ -1160,7 +1230,7 @@ TEST_P(Test_DLDT_two_inputs, as_backend)
INSTANTIATE_TEST_CASE_P(/*nothing*/, Test_DLDT_two_inputs, Combine(
Values(CV_8U, CV_32F), Values(CV_8U, CV_32F),
testing::ValuesIn(getAvailableTargets(DNN_BACKEND_INFERENCE_ENGINE))
dnnBackendsAndTargets()
));
class UnsupportedLayer : public Layer
@@ -1181,10 +1251,11 @@ public:
virtual void forward(cv::InputArrayOfArrays inputs, cv::OutputArrayOfArrays outputs, cv::OutputArrayOfArrays internals) CV_OVERRIDE {}
};
TEST(Test_DLDT, fused_output)
typedef DNNTestLayer Test_DLDT_layers;
static void test_dldt_fused_output(Backend backend, Target target)
{
static const int kNumChannels = 3;
CV_DNN_REGISTER_LAYER_CLASS(Unsupported, UnsupportedLayer);
Net net;
{
LayerParams lp;
@@ -1208,13 +1279,31 @@ TEST(Test_DLDT, fused_output)
LayerParams lp;
net.addLayerToPrev("unsupported_layer", "Unsupported", lp);
}
net.setPreferableBackend(DNN_BACKEND_INFERENCE_ENGINE);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(Mat({1, 1, 1, 1}, CV_32FC1, Scalar(1)));
ASSERT_NO_THROW(net.forward());
net.forward();
}
TEST_P(Test_DLDT_layers, fused_output)
{
CV_DNN_REGISTER_LAYER_CLASS(Unsupported, UnsupportedLayer);
try
{
test_dldt_fused_output(backend, target);
}
catch (const std::exception& e)
{
ADD_FAILURE() << "Exception: " << e.what();
}
catch(...)
{
ADD_FAILURE() << "Unknown exception";
}
LayerFactory::unregisterLayer("Unsupported");
}
TEST(Test_DLDT, multiple_networks)
TEST_P(Test_DLDT_layers, multiple_networks)
{
Net nets[2];
for (int i = 0; i < 2; ++i)
@@ -1229,7 +1318,8 @@ TEST(Test_DLDT, multiple_networks)
lp.name = format("testConv_%d", i);
lp.blobs.push_back(Mat({1, 1, 1, 1}, CV_32F, Scalar(1 + i)));
nets[i].addLayerToPrev(lp.name, lp.type, lp);
nets[i].setPreferableBackend(DNN_BACKEND_INFERENCE_ENGINE);
nets[i].setPreferableBackend(backend);
nets[i].setPreferableTarget(target);
nets[i].setInput(Mat({1, 1, 1, 1}, CV_32FC1, Scalar(1)));
}
Mat out_1 = nets[0].forward();
@@ -1238,6 +1328,9 @@ TEST(Test_DLDT, multiple_networks)
out_1 = nets[0].forward();
normAssert(2 * out_1, out_2);
}
INSTANTIATE_TEST_CASE_P(/*nothing*/, Test_DLDT_layers, dnnBackendsAndTargets());
#endif // HAVE_INF_ENGINE
// Test a custom layer.
@@ -1353,7 +1446,7 @@ TEST_P(Test_Caffe_layers, Interp)
TEST_P(Test_Caffe_layers, DISABLED_Interp) // requires patched protobuf (available in OpenCV source tree only)
#endif
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
if (backend == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 && target == DNN_TARGET_MYRIAD)
applyTestTag(CV_TEST_TAG_DNN_SKIP_IE_MYRIAD);
// Test a custom layer.