static void runTorchNet(String prefix, int targetId = DNN_TARGET_CPU, String outLayerName = "",
bool check2ndBlob = false, bool isBinary = false)
{
String suffix = (isBinary) ? ".dat" : ".txt";
Net net = readNetFromTorch(_tf(prefix + "_net" + suffix), isBinary);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(DNN_BACKEND_DEFAULT);
net.setPreferableTarget(targetId);
Mat inp, outRef;
ASSERT_NO_THROW( inp = readTorchBlob(_tf(prefix + "_input" + suffix), isBinary) );
ASSERT_NO_THROW( outRef = readTorchBlob(_tf(prefix + "_output" + suffix), isBinary) );
if (outLayerName.empty())
outLayerName = net.getLayerNames().back();
net.setInput(inp, "0");
std::vector<Mat> outBlobs;
net.forward(outBlobs, outLayerName);
normAssert(outRef, outBlobs[0]);
if (check2ndBlob)
{
Mat out2 = outBlobs[1];
Mat ref2 = readTorchBlob(_tf(prefix + "_output_2" + suffix), isBinary);
normAssert(out2, ref2);
}
}
TEST_P(Test_TensorFlow_nets, opencv_face_detector_uint8)
{
checkBackend();
std::string proto = findDataFile("dnn/opencv_face_detector.pbtxt", false);
std::string model = findDataFile("dnn/opencv_face_detector_uint8.pb", false);
Net net = readNetFromTensorflow(model, proto);
Mat img = imread(findDataFile("gpu/lbpcascade/er.png", false));
Mat blob = blobFromImage(img, 1.0, Size(), Scalar(104.0, 177.0, 123.0), false, false);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(blob);
// Output has shape 1x1xNx7 where N - number of detections.
// An every detection is a vector of values [id, classId, confidence, left, top, right, bottom]
Mat out = net.forward();
// References are from test for Caffe model.
Mat ref = (Mat_<float>(6, 7) << 0, 1, 0.99520785, 0.80997437, 0.16379407, 0.87996572, 0.26685631,
0, 1, 0.9934696, 0.2831718, 0.50738752, 0.345781, 0.5985168,
0, 1, 0.99096733, 0.13629119, 0.24892329, 0.19756334, 0.3310290,
0, 1, 0.98977017, 0.23901358, 0.09084064, 0.29902688, 0.1769477,
0, 1, 0.97203469, 0.67965847, 0.06876482, 0.73999709, 0.1513494,
0, 1, 0.95097077, 0.51901293, 0.45863652, 0.5777427, 0.5347801);
double scoreDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 4e-3 : 3.4e-3;
double iouDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 0.024 : 1e-2;
normAssertDetections(ref, out, "", 0.9, scoreDiff, iouDiff);
}
TEST_P(Test_TensorFlow_nets, Inception_v2_SSD)
{
checkBackend();
std::string proto = findDataFile("dnn/ssd_inception_v2_coco_2017_11_17.pbtxt", false);
std::string model = findDataFile("dnn/ssd_inception_v2_coco_2017_11_17.pb", false);
Net net = readNetFromTensorflow(model, proto);
Mat img = imread(findDataFile("dnn/street.png", false));
Mat blob = blobFromImage(img, 1.0f, Size(300, 300), Scalar(), true, false);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(blob);
// Output has shape 1x1xNx7 where N - number of detections.
// An every detection is a vector of values [id, classId, confidence, left, top, right, bottom]
Mat out = net.forward();
Mat ref = (Mat_<float>(5, 7) << 0, 1, 0.90176028, 0.19872092, 0.36311883, 0.26461923, 0.63498729,
0, 3, 0.93569964, 0.64865261, 0.45906419, 0.80675775, 0.65708131,
0, 3, 0.75838411, 0.44668293, 0.45907149, 0.49459291, 0.52197015,
0, 10, 0.95932811, 0.38349164, 0.32528657, 0.40387636, 0.39165527,
0, 10, 0.93973452, 0.66561931, 0.37841269, 0.68074018, 0.42907384);
double scoreDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 0.0097 : default_l1;
double iouDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 0.09 : default_lInf;
normAssertDetections(ref, out, "", 0.5, scoreDiff, iouDiff);
}
TEST_P(Test_TensorFlow_nets, Faster_RCNN)
{
static std::string names[] = {"faster_rcnn_inception_v2_coco_2018_01_28",
"faster_rcnn_resnet50_coco_2018_01_28"};
checkBackend();
if ((backend == DNN_BACKEND_INFERENCE_ENGINE && target != DNN_TARGET_CPU) ||
(backend == DNN_BACKEND_OPENCV && target == DNN_TARGET_OPENCL_FP16))
throw SkipTestException("");
for (int i = 1; i < 2; ++i)
{
std::string proto = findDataFile("dnn/" + names[i] + ".pbtxt", false);
std::string model = findDataFile("dnn/" + names[i] + ".pb", false);
Net net = readNetFromTensorflow(model, proto);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat img = imread(findDataFile("dnn/dog416.png", false));
Mat blob = blobFromImage(img, 1.0f, Size(800, 600), Scalar(), true, false);
net.setInput(blob);
Mat out = net.forward();
Mat ref = blobFromNPY(findDataFile("dnn/tensorflow/" + names[i] + ".detection_out.npy"));
normAssertDetections(ref, out, names[i].c_str(), 0.3);
}
}
TEST_P(Test_TensorFlow_nets, Inception_v2_SSD)
{
std::string proto = findDataFile("dnn/ssd_inception_v2_coco_2017_11_17.pbtxt", false);
std::string model = findDataFile("dnn/ssd_inception_v2_coco_2017_11_17.pb", false);
Net net = readNetFromTensorflow(model, proto);
Mat img = imread(findDataFile("dnn/street.png", false));
Mat blob = blobFromImage(img, 1.0f / 127.5, Size(300, 300), Scalar(127.5, 127.5, 127.5), true, false);
net.setPreferableTarget(GetParam());
net.setInput(blob);
// Output has shape 1x1xNx7 where N - number of detections.
// An every detection is a vector of values [id, classId, confidence, left, top, right, bottom]
Mat out = net.forward();
out = out.reshape(1, out.total() / 7);
Mat detections;
for (int i = 0; i < out.rows; ++i)
{
if (out.at<float>(i, 2) > 0.5)
detections.push_back(out.row(i).colRange(1, 7));
}
Mat ref = (Mat_<float>(5, 6) << 1, 0.90176028, 0.19872092, 0.36311883, 0.26461923, 0.63498729,
3, 0.93569964, 0.64865261, 0.45906419, 0.80675775, 0.65708131,
3, 0.75838411, 0.44668293, 0.45907149, 0.49459291, 0.52197015,
10, 0.95932811, 0.38349164, 0.32528657, 0.40387636, 0.39165527,
10, 0.93973452, 0.66561931, 0.37841269, 0.68074018, 0.42907384);
normAssert(detections, ref);
}
TEST_P(Test_TensorFlow_nets, MobileNet_SSD)
{
std::string netPath = findDataFile("dnn/ssd_mobilenet_v1_coco.pb", false);
std::string netConfig = findDataFile("dnn/ssd_mobilenet_v1_coco.pbtxt", false);
std::string imgPath = findDataFile("dnn/street.png", false);
Mat inp;
resize(imread(imgPath), inp, Size(300, 300));
inp = blobFromImage(inp, 1.0f / 127.5, Size(), Scalar(127.5, 127.5, 127.5), true);
std::vector<String> outNames(3);
outNames[0] = "concat";
outNames[1] = "concat_1";
outNames[2] = "detection_out";
std::vector<Mat> target(outNames.size());
for (int i = 0; i < outNames.size(); ++i)
{
std::string path = findDataFile("dnn/tensorflow/ssd_mobilenet_v1_coco." + outNames[i] + ".npy", false);
target[i] = blobFromNPY(path);
}
Net net = readNetFromTensorflow(netPath, netConfig);
net.setPreferableTarget(GetParam());
net.setInput(inp);
std::vector<Mat> output;
net.forward(output, outNames);
normAssert(target[0].reshape(1, 1), output[0].reshape(1, 1), "", 1e-5, 1.5e-4);
normAssert(target[1].reshape(1, 1), output[1].reshape(1, 1), "", 1e-5, 3e-4);
normAssert(target[2].reshape(1, 1), output[2].reshape(1, 1), "", 4e-5, 1e-2);
}
// Check accuracy of style transfer models from https://github.com/jcjohnson/fast-neural-style
// th fast_neural_style.lua \
// -input_image ~/opencv_extra/testdata/dnn/googlenet_1.png \
// -output_image lena.png \
// -median_filter 0 \
// -image_size 0 \
// -model models/eccv16/starry_night.t7
// th fast_neural_style.lua \
// -input_image ~/opencv_extra/testdata/dnn/googlenet_1.png \
// -output_image lena.png \
// -median_filter 0 \
// -image_size 0 \
// -model models/instance_norm/feathers.t7
TEST_P(Test_Torch_nets, FastNeuralStyle_accuracy)
{
std::string models[] = {"dnn/fast_neural_style_eccv16_starry_night.t7",
"dnn/fast_neural_style_instance_norm_feathers.t7"};
std::string targets[] = {"dnn/lena_starry_night.png", "dnn/lena_feathers.png"};
for (int i = 0; i < 2; ++i)
{
const string model = findDataFile(models[i], false);
Net net = readNetFromTorch(model);
net.setPreferableTarget(GetParam());
Mat img = imread(findDataFile("dnn/googlenet_1.png", false));
Mat inputBlob = blobFromImage(img, 1.0, Size(), Scalar(103.939, 116.779, 123.68), false);
net.setInput(inputBlob);
Mat out = net.forward();
// Deprocessing.
getPlane(out, 0, 0) += 103.939;
getPlane(out, 0, 1) += 116.779;
getPlane(out, 0, 2) += 123.68;
out = cv::min(cv::max(0, out), 255);
Mat ref = imread(findDataFile(targets[i]));
Mat refBlob = blobFromImage(ref, 1.0, Size(), Scalar(), false);
normAssert(out, refBlob, "", 0.5, 1.1);
}
}
void runTorchNet(const String& prefix, String outLayerName = "",
bool check2ndBlob = false, bool isBinary = false,
double l1 = 0.0, double lInf = 0.0)
{
String suffix = (isBinary) ? ".dat" : ".txt";
Mat inp, outRef;
ASSERT_NO_THROW( inp = readTorchBlob(_tf(prefix + "_input" + suffix), isBinary) );
ASSERT_NO_THROW( outRef = readTorchBlob(_tf(prefix + "_output" + suffix), isBinary) );
checkBackend(backend, target, &inp, &outRef);
Net net = readNetFromTorch(_tf(prefix + "_net" + suffix), isBinary);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
if (outLayerName.empty())
outLayerName = net.getLayerNames().back();
net.setInput(inp);
std::vector<Mat> outBlobs;
net.forward(outBlobs, outLayerName);
l1 = l1 ? l1 : default_l1;
lInf = lInf ? lInf : default_lInf;
normAssert(outRef, outBlobs[0], "", l1, lInf);
if (check2ndBlob && backend != DNN_BACKEND_INFERENCE_ENGINE)
{
Mat out2 = outBlobs[1];
Mat ref2 = readTorchBlob(_tf(prefix + "_output_2" + suffix), isBinary);
normAssert(out2, ref2, "", l1, lInf);
}
}
TEST_P(Test_TensorFlow_nets, opencv_face_detector_uint8)
{
std::string proto = findDataFile("dnn/opencv_face_detector.pbtxt", false);
std::string model = findDataFile("dnn/opencv_face_detector_uint8.pb", false);
Net net = readNetFromTensorflow(model, proto);
Mat img = imread(findDataFile("gpu/lbpcascade/er.png", false));
Mat blob = blobFromImage(img, 1.0, Size(), Scalar(104.0, 177.0, 123.0), false, false);
net.setPreferableTarget(GetParam());
net.setInput(blob);
// Output has shape 1x1xNx7 where N - number of detections.
// An every detection is a vector of values [id, classId, confidence, left, top, right, bottom]
Mat out = net.forward();
// References are from test for Caffe model.
Mat ref = (Mat_<float>(6, 5) << 0.99520785, 0.80997437, 0.16379407, 0.87996572, 0.26685631,
0.9934696, 0.2831718, 0.50738752, 0.345781, 0.5985168,
0.99096733, 0.13629119, 0.24892329, 0.19756334, 0.3310290,
0.98977017, 0.23901358, 0.09084064, 0.29902688, 0.1769477,
0.97203469, 0.67965847, 0.06876482, 0.73999709, 0.1513494,
0.95097077, 0.51901293, 0.45863652, 0.5777427, 0.5347801);
normAssert(out.reshape(1, out.total() / 7).rowRange(0, 6).colRange(2, 7), ref, "", 2.8e-4, 3.4e-3);
}
TEST_P(Test_Torch_nets, ENet_accuracy)
{
Net net;
{
const string model = findDataFile("dnn/Enet-model-best.net", false);
net = readNetFromTorch(model, true);
ASSERT_TRUE(!net.empty());
}
net.setPreferableTarget(GetParam());
Mat sample = imread(_tf("street.png", false));
Mat inputBlob = blobFromImage(sample, 1./255);
net.setInput(inputBlob, "");
Mat out = net.forward();
Mat ref = blobFromNPY(_tf("torch_enet_prob.npy", false));
// Due to numerical instability in Pooling-Unpooling layers (indexes jittering)
// thresholds for ENet must be changed. Accuracy of results was checked on
// Cityscapes dataset and difference in mIOU with Torch is 10E-4%
normAssert(ref, out, "", 0.00044, 0.44);
const int N = 3;
for (int i = 0; i < N; i++)
{
net.setInput(inputBlob, "");
Mat out = net.forward();
normAssert(ref, out, "", 0.00044, 0.44);
}
}
TEST_P(opencv_face_detector, Accuracy)
{
std::string proto = findDataFile("dnn/opencv_face_detector.prototxt", false);
std::string model = findDataFile(get<0>(GetParam()), false);
dnn::Target targetId = (dnn::Target)(int)get<1>(GetParam());
Net net = readNetFromCaffe(proto, model);
Mat img = imread(findDataFile("gpu/lbpcascade/er.png", false));
Mat blob = blobFromImage(img, 1.0, Size(), Scalar(104.0, 177.0, 123.0), false, false);
net.setPreferableBackend(DNN_BACKEND_DEFAULT);
net.setPreferableTarget(targetId);
net.setInput(blob);
// Output has shape 1x1xNx7 where N - number of detections.
// An every detection is a vector of values [id, classId, confidence, left, top, right, bottom]
Mat out = net.forward();
Mat ref = (Mat_<float>(6, 7) << 0, 1, 0.99520785, 0.80997437, 0.16379407, 0.87996572, 0.26685631,
0, 1, 0.9934696, 0.2831718, 0.50738752, 0.345781, 0.5985168,
0, 1, 0.99096733, 0.13629119, 0.24892329, 0.19756334, 0.3310290,
0, 1, 0.98977017, 0.23901358, 0.09084064, 0.29902688, 0.1769477,
0, 1, 0.97203469, 0.67965847, 0.06876482, 0.73999709, 0.1513494,
0, 1, 0.95097077, 0.51901293, 0.45863652, 0.5777427, 0.5347801);
normAssertDetections(ref, out, "", 0.5, 1e-5, 2e-4);
}
TEST_P(Reproducibility_SqueezeNet_v1_1, Accuracy)
{
Net net = readNetFromCaffe(findDataFile("dnn/squeezenet_v1.1.prototxt", false),
findDataFile("dnn/squeezenet_v1.1.caffemodel", false));
int targetId = GetParam();
net.setPreferableTarget(targetId);
Mat input = blobFromImage(imread(_tf("googlenet_0.png")), 1.0f, Size(227,227), Scalar(), false);
ASSERT_TRUE(!input.empty());
Mat out;
if (targetId == DNN_TARGET_OPENCL)
{
// Firstly set a wrong input blob and run the model to receive a wrong output.
// Then set a correct input blob to check CPU->GPU synchronization is working well.
net.setInput(input * 2.0f);
out = net.forward();
}
net.setInput(input);
out = net.forward();
Mat ref = blobFromNPY(_tf("squeezenet_v1.1_prob.npy"));
normAssert(ref, out);
}
TEST_P(Reproducibility_ResNet50, Accuracy)
{
Net net = readNetFromCaffe(findDataFile("dnn/ResNet-50-deploy.prototxt", false),
findDataFile("dnn/ResNet-50-model.caffemodel", false));
int targetId = GetParam();
net.setPreferableTarget(targetId);
float l1 = (targetId == DNN_TARGET_OPENCL_FP16) ? 3e-5 : 1e-5;
float lInf = (targetId == DNN_TARGET_OPENCL_FP16) ? 6e-3 : 1e-4;
Mat input = blobFromImage(imread(_tf("googlenet_0.png")), 1.0f, Size(224,224), Scalar(), false);
ASSERT_TRUE(!input.empty());
net.setInput(input);
Mat out = net.forward();
Mat ref = blobFromNPY(_tf("resnet50_prob.npy"));
normAssert(ref, out, "", l1, lInf);
if (targetId == DNN_TARGET_OPENCL || targetId == DNN_TARGET_OPENCL_FP16)
{
UMat out_umat;
net.forward(out_umat);
normAssert(ref, out_umat, "out_umat", l1, lInf);
std::vector<UMat> out_umats;
net.forward(out_umats);
normAssert(ref, out_umats[0], "out_umat_vector", l1, lInf);
}
}
OCL_TEST(Torch_Importer, ENet_accuracy)
{
Net net;
{
const string model = findDataFile("dnn/Enet-model-best.net", false);
Ptr<Importer> importer = createTorchImporter(model, true);
ASSERT_TRUE(importer != NULL);
importer->populateNet(net);
}
net.setPreferableBackend(DNN_BACKEND_DEFAULT);
net.setPreferableTarget(DNN_TARGET_OPENCL);
Mat sample = imread(_tf("street.png", false));
Mat inputBlob = blobFromImage(sample, 1./255);
net.setInput(inputBlob, "");
Mat out = net.forward();
Mat ref = blobFromNPY(_tf("torch_enet_prob.npy", false));
// Due to numerical instability in Pooling-Unpooling layers (indexes jittering)
// thresholds for ENet must be changed. Accuracy of resuults was checked on
// Cityscapes dataset and difference in mIOU with Torch is 10E-4%
normAssert(ref, out, "", 0.00044, 0.44);
const int N = 3;
for (int i = 0; i < N; i++)
{
net.setInput(inputBlob, "");
Mat out = net.forward();
normAssert(ref, out, "", 0.00044, 0.44);
}
}
TEST_P(Test_Torch_nets, OpenFace_accuracy)
{
#if defined(INF_ENGINE_RELEASE) && INF_ENGINE_RELEASE < 2018030000
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
throw SkipTestException("Test is enabled starts from OpenVINO 2018R3");
#endif
checkBackend();
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_OPENCL_FP16)
throw SkipTestException("");
const string model = findDataFile("dnn/openface_nn4.small2.v1.t7", false);
Net net = readNetFromTorch(model);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat sample = imread(findDataFile("cv/shared/lena.png", false));
Mat sampleF32(sample.size(), CV_32FC3);
sample.convertTo(sampleF32, sampleF32.type());
sampleF32 /= 255;
resize(sampleF32, sampleF32, Size(96, 96), 0, 0, INTER_NEAREST);
Mat inputBlob = blobFromImage(sampleF32, 1.0, Size(), Scalar(), /*swapRB*/true);
net.setInput(inputBlob);
Mat out = net.forward();
Mat outRef = readTorchBlob(_tf("net_openface_output.dat"), true);
normAssert(out, outRef, "", default_l1, default_lInf);
}
OCL_TEST(Reproducibility_TinyYoloVoc, Accuracy)
{
Net net;
{
const string cfg = findDataFile("dnn/tiny-yolo-voc.cfg", false);
const string model = findDataFile("dnn/tiny-yolo-voc.weights", false);
net = readNetFromDarknet(cfg, model);
ASSERT_FALSE(net.empty());
}
net.setPreferableBackend(DNN_BACKEND_DEFAULT);
net.setPreferableTarget(DNN_TARGET_OPENCL);
// dog416.png is dog.jpg that resized to 416x416 in the lossless PNG format
Mat sample = imread(_tf("dog416.png"));
ASSERT_TRUE(!sample.empty());
Size inputSize(416, 416);
if (sample.size() != inputSize)
resize(sample, sample, inputSize);
net.setInput(blobFromImage(sample, 1 / 255.F), "data");
Mat out = net.forward("detection_out");
Mat detection;
const float confidenceThreshold = 0.24;
for (int i = 0; i < out.rows; i++) {
const int probability_index = 5;
const int probability_size = out.cols - probability_index;
float *prob_array_ptr = &out.at<float>(i, probability_index);
size_t objectClass = std::max_element(prob_array_ptr, prob_array_ptr + probability_size) - prob_array_ptr;
float confidence = out.at<float>(i, (int)objectClass + probability_index);
if (confidence > confidenceThreshold)
detection.push_back(out.row(i));
}
// obtained by: ./darknet detector test ./cfg/voc.data ./cfg/tiny-yolo-voc.cfg ./tiny-yolo-voc.weights -thresh 0.24 ./dog416.png
// There are 2 objects (6-car, 11-dog) with 25 values for each:
// { relative_center_x, relative_center_y, relative_width, relative_height, unused_t0, probability_for_each_class[20] }
float ref_array[] = {
0.736762F, 0.239551F, 0.315440F, 0.160779F, 0.761977F, 0.000000F, 0.000000F, 0.000000F, 0.000000F,
0.000000F, 0.000000F, 0.761967F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F,
0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F,
0.287486F, 0.653731F, 0.315579F, 0.534527F, 0.782737F, 0.000000F, 0.000000F, 0.000000F, 0.000000F,
0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.780595F,
0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F
};
const int number_of_objects = 2;
Mat ref(number_of_objects, sizeof(ref_array) / (number_of_objects * sizeof(float)), CV_32FC1, &ref_array);
normAssert(ref, detection);
}
// inp = cv.imread('opencv_extra/testdata/cv/ximgproc/sources/08.png')
// inp = inp[:,:,[2, 1, 0]].astype(np.float32).reshape(1, 512, 512, 3)
// outs = sess.run([sess.graph.get_tensor_by_name('feature_fusion/Conv_7/Sigmoid:0'),
// sess.graph.get_tensor_by_name('feature_fusion/concat_3:0')],
// feed_dict={'input_images:0': inp})
// scores = np.ascontiguousarray(outs[0].transpose(0, 3, 1, 2))
// geometry = np.ascontiguousarray(outs[1].transpose(0, 3, 1, 2))
// np.save('east_text_detection.scores.npy', scores)
// np.save('east_text_detection.geometry.npy', geometry)
TEST_P(Test_TensorFlow_nets, EAST_text_detection)
{
checkBackend();
#if defined(INF_ENGINE_RELEASE) && INF_ENGINE_RELEASE < 2018030000
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_MYRIAD)
throw SkipTestException("Test is enabled starts from OpenVINO 2018R3");
#endif
std::string netPath = findDataFile("dnn/frozen_east_text_detection.pb", false);
std::string imgPath = findDataFile("cv/ximgproc/sources/08.png", false);
std::string refScoresPath = findDataFile("dnn/east_text_detection.scores.npy", false);
std::string refGeometryPath = findDataFile("dnn/east_text_detection.geometry.npy", false);
Net net = readNet(findDataFile("dnn/frozen_east_text_detection.pb", false));
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat img = imread(imgPath);
Mat inp = blobFromImage(img, 1.0, Size(), Scalar(123.68, 116.78, 103.94), true, false);
net.setInput(inp);
std::vector<Mat> outs;
std::vector<String> outNames(2);
outNames[0] = "feature_fusion/Conv_7/Sigmoid";
outNames[1] = "feature_fusion/concat_3";
net.forward(outs, outNames);
Mat scores = outs[0];
Mat geometry = outs[1];
// Scores are in range [0, 1]. Geometry values are in range [-0.23, 290]
double l1_scores = default_l1, lInf_scores = default_lInf;
double l1_geometry = default_l1, lInf_geometry = default_lInf;
if (target == DNN_TARGET_OPENCL_FP16)
{
lInf_scores = 0.11;
l1_geometry = 0.28; lInf_geometry = 5.94;
}
else if (target == DNN_TARGET_MYRIAD)
{
lInf_scores = 0.214;
l1_geometry = 0.47; lInf_geometry = 15.34;
}
else
{
l1_geometry = 1e-4, lInf_geometry = 3e-3;
}
normAssert(scores, blobFromNPY(refScoresPath), "scores", l1_scores, lInf_scores);
normAssert(geometry, blobFromNPY(refGeometryPath), "geometry", l1_geometry, lInf_geometry);
}
// Check accuracy of style transfer models from https://github.com/jcjohnson/fast-neural-style
// th fast_neural_style.lua \
// -input_image ~/opencv_extra/testdata/dnn/googlenet_1.png \
// -output_image lena.png \
// -median_filter 0 \
// -image_size 0 \
// -model models/eccv16/starry_night.t7
// th fast_neural_style.lua \
// -input_image ~/opencv_extra/testdata/dnn/googlenet_1.png \
// -output_image lena.png \
// -median_filter 0 \
// -image_size 0 \
// -model models/instance_norm/feathers.t7
TEST_P(Test_Torch_nets, FastNeuralStyle_accuracy)
{
checkBackend();
#if defined(INF_ENGINE_RELEASE)
#if INF_ENGINE_RELEASE <= 2018050000
if (backend == DNN_BACKEND_INFERENCE_ENGINE && target == DNN_TARGET_OPENCL)
throw SkipTestException("");
#endif
#endif
std::string models[] = {"dnn/fast_neural_style_eccv16_starry_night.t7",
"dnn/fast_neural_style_instance_norm_feathers.t7"};
std::string targets[] = {"dnn/lena_starry_night.png", "dnn/lena_feathers.png"};
for (int i = 0; i < 2; ++i)
{
const string model = findDataFile(models[i], false);
Net net = readNetFromTorch(model);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat img = imread(findDataFile("dnn/googlenet_1.png", false));
Mat inputBlob = blobFromImage(img, 1.0, Size(), Scalar(103.939, 116.779, 123.68), false);
net.setInput(inputBlob);
Mat out = net.forward();
// Deprocessing.
getPlane(out, 0, 0) += 103.939;
getPlane(out, 0, 1) += 116.779;
getPlane(out, 0, 2) += 123.68;
out = cv::min(cv::max(0, out), 255);
Mat ref = imread(findDataFile(targets[i]));
Mat refBlob = blobFromImage(ref, 1.0, Size(), Scalar(), false);
if (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD)
{
double normL1 = cvtest::norm(refBlob, out, cv::NORM_L1) / refBlob.total();
if (target == DNN_TARGET_MYRIAD)
EXPECT_LE(normL1, 4.0f);
else
EXPECT_LE(normL1, 0.6f);
}
else
normAssert(out, refBlob, "", 0.5, 1.1);
}
}
void testDarknetLayer(const std::string& name, bool hasWeights = false)
{
std::string cfg = findDataFile("dnn/darknet/" + name + ".cfg", false);
std::string model = "";
if (hasWeights)
model = findDataFile("dnn/darknet/" + name + ".weights", false);
Mat inp = blobFromNPY(findDataFile("dnn/darknet/" + name + "_in.npy", false));
Mat ref = blobFromNPY(findDataFile("dnn/darknet/" + name + "_out.npy", false));
checkBackend(&inp, &ref);
Net net = readNet(cfg, model);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(inp);
Mat out = net.forward();
normAssert(out, ref, "", default_l1, default_lInf);
}
void testONNXModels(const String& basename, const Extension ext = npy,
const double l1 = 0, const float lInf = 0, const bool useSoftmax = false,
bool checkNoFallbacks = true)
{
String onnxmodel = _tf("models/" + basename + ".onnx");
Mat inp, ref;
if (ext == npy) {
inp = blobFromNPY(_tf("data/input_" + basename + ".npy"));
ref = blobFromNPY(_tf("data/output_" + basename + ".npy"));
}
else if (ext == pb) {
inp = readTensorFromONNX(_tf("data/input_" + basename + ".pb"));
ref = readTensorFromONNX(_tf("data/output_" + basename + ".pb"));
}
else
CV_Error(Error::StsUnsupportedFormat, "Unsupported extension");
checkBackend(&inp, &ref);
Net net = readNetFromONNX(onnxmodel);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(inp);
Mat out = net.forward("");
if (useSoftmax)
{
LayerParams lp;
Net netSoftmax;
netSoftmax.addLayerToPrev("softmaxLayer", "SoftMax", lp);
netSoftmax.setPreferableBackend(DNN_BACKEND_OPENCV);
netSoftmax.setInput(out);
out = netSoftmax.forward();
netSoftmax.setInput(ref);
ref = netSoftmax.forward();
}
normAssert(ref, out, "", l1 ? l1 : default_l1, lInf ? lInf : default_lInf);
if (checkNoFallbacks)
expectNoFallbacksFromIE(net);
}
TEST_P(Test_Torch_nets, OpenFace_accuracy)
{
const string model = findDataFile("dnn/openface_nn4.small2.v1.t7", false);
Net net = readNetFromTorch(model);
net.setPreferableTarget(GetParam());
Mat sample = imread(findDataFile("cv/shared/lena.png", false));
Mat sampleF32(sample.size(), CV_32FC3);
sample.convertTo(sampleF32, sampleF32.type());
sampleF32 /= 255;
resize(sampleF32, sampleF32, Size(96, 96), 0, 0, INTER_NEAREST);
Mat inputBlob = blobFromImage(sampleF32);
net.setInput(inputBlob);
Mat out = net.forward();
Mat outRef = readTorchBlob(_tf("net_openface_output.dat"), true);
normAssert(out, outRef);
}
TEST_P(Test_TensorFlow_nets, MobileNet_v1_SSD_PPN)
{
checkBackend();
std::string proto = findDataFile("dnn/ssd_mobilenet_v1_ppn_coco.pbtxt", false);
std::string model = findDataFile("dnn/ssd_mobilenet_v1_ppn_coco.pb", false);
Net net = readNetFromTensorflow(model, proto);
Mat img = imread(findDataFile("dnn/dog416.png", false));
Mat ref = blobFromNPY(findDataFile("dnn/tensorflow/ssd_mobilenet_v1_ppn_coco.detection_out.npy", false));
Mat blob = blobFromImage(img, 1.0f, Size(300, 300), Scalar(), true, false);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(blob);
Mat out = net.forward();
double scoreDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 0.011 : default_l1;
double iouDiff = (target == DNN_TARGET_OPENCL_FP16 || target == DNN_TARGET_MYRIAD) ? 0.021 : default_lInf;
normAssertDetections(ref, out, "", 0.4, scoreDiff, iouDiff);
}
TEST_P(Test_TensorFlow_nets, MobileNet_SSD)
{
checkBackend();
if ((backend == DNN_BACKEND_INFERENCE_ENGINE && target != DNN_TARGET_CPU) ||
(backend == DNN_BACKEND_OPENCV && target == DNN_TARGET_OPENCL_FP16))
throw SkipTestException("");
std::string netPath = findDataFile("dnn/ssd_mobilenet_v1_coco.pb", false);
std::string netConfig = findDataFile("dnn/ssd_mobilenet_v1_coco.pbtxt", false);
std::string imgPath = findDataFile("dnn/street.png", false);
Mat inp;
resize(imread(imgPath), inp, Size(300, 300));
inp = blobFromImage(inp, 1.0f / 127.5, Size(), Scalar(127.5, 127.5, 127.5), true);
std::vector<String> outNames(3);
outNames[0] = "concat";
outNames[1] = "concat_1";
outNames[2] = "detection_out";
std::vector<Mat> refs(outNames.size());
for (int i = 0; i < outNames.size(); ++i)
{
std::string path = findDataFile("dnn/tensorflow/ssd_mobilenet_v1_coco." + outNames[i] + ".npy", false);
refs[i] = blobFromNPY(path);
}
Net net = readNetFromTensorflow(netPath, netConfig);
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(inp);
std::vector<Mat> output;
net.forward(output, outNames);
normAssert(refs[0].reshape(1, 1), output[0].reshape(1, 1), "", 1e-5, 1.5e-4);
normAssert(refs[1].reshape(1, 1), output[1].reshape(1, 1), "", 1e-5, 3e-4);
normAssertDetections(refs[2], output[2], "", 0.2);
}
void runTensorFlowNet(const std::string& prefix, bool hasText = false,
double l1 = 0.0, double lInf = 0.0, bool memoryLoad = false)
{
std::string netPath = path(prefix + "_net.pb");
std::string netConfig = (hasText ? path(prefix + "_net.pbtxt") : "");
std::string inpPath = path(prefix + "_in.npy");
std::string outPath = path(prefix + "_out.npy");
cv::Mat input = blobFromNPY(inpPath);
cv::Mat ref = blobFromNPY(outPath);
checkBackend(&input, &ref);
Net net;
if (memoryLoad)
{
// Load files into a memory buffers
string dataModel;
ASSERT_TRUE(readFileInMemory(netPath, dataModel));
string dataConfig;
if (hasText)
{
ASSERT_TRUE(readFileInMemory(netConfig, dataConfig));
}
net = readNetFromTensorflow(dataModel.c_str(), dataModel.size(),
dataConfig.c_str(), dataConfig.size());
}
else
net = readNetFromTensorflow(netPath, netConfig);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
net.setInput(input);
cv::Mat output = net.forward();
normAssert(ref, output, "", l1 ? l1 : default_l1, lInf ? lInf : default_lInf);
}
TEST_P(Test_ONNX_nets, Alexnet)
{
applyTestTag(target == DNN_TARGET_CPU ? CV_TEST_TAG_MEMORY_512MB : CV_TEST_TAG_MEMORY_1GB);
const String model = _tf("models/alexnet.onnx");
Net net = readNetFromONNX(model);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat inp = imread(_tf("../grace_hopper_227.png"));
Mat ref = blobFromNPY(_tf("../caffe_alexnet_prob.npy"));
checkBackend(&inp, &ref);
net.setInput(blobFromImage(inp, 1.0f, Size(227, 227), Scalar(), false));
ASSERT_FALSE(net.empty());
Mat out = net.forward();
normAssert(out, ref, "", default_l1, default_lInf);
expectNoFallbacksFromIE(net);
}
TEST_P(Test_ONNX_layers, MultyInputs)
{
const String model = _tf("models/multy_inputs.onnx");
Net net = readNetFromONNX(model);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat inp1 = blobFromNPY(_tf("data/input_multy_inputs_0.npy"));
Mat inp2 = blobFromNPY(_tf("data/input_multy_inputs_1.npy"));
Mat ref = blobFromNPY(_tf("data/output_multy_inputs.npy"));
checkBackend(&inp1, &ref);
net.setInput(inp1, "0");
net.setInput(inp2, "1");
Mat out = net.forward();
normAssert(ref, out, "", default_l1, default_lInf);
expectNoFallbacksFromIE(net);
}
static void runTensorFlowNet(const std::string& prefix, int targetId = DNN_TARGET_CPU, bool hasText = false,
double l1 = 1e-5, double lInf = 1e-4,
bool memoryLoad = false)
{
std::string netPath = path(prefix + "_net.pb");
std::string netConfig = (hasText ? path(prefix + "_net.pbtxt") : "");
std::string inpPath = path(prefix + "_in.npy");
std::string outPath = path(prefix + "_out.npy");
Net net;
if (memoryLoad)
{
// Load files into a memory buffers
string dataModel;
ASSERT_TRUE(readFileInMemory(netPath, dataModel));
string dataConfig;
if (hasText)
ASSERT_TRUE(readFileInMemory(netConfig, dataConfig));
net = readNetFromTensorflow(dataModel.c_str(), dataModel.size(),
dataConfig.c_str(), dataConfig.size());
}
else
net = readNetFromTensorflow(netPath, netConfig);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(DNN_BACKEND_DEFAULT);
net.setPreferableTarget(targetId);
cv::Mat input = blobFromNPY(inpPath);
cv::Mat target = blobFromNPY(outPath);
net.setInput(input);
cv::Mat output = net.forward();
normAssert(target, output, "", l1, lInf);
}
TEST_P(Test_Torch_nets, ENet_accuracy)
{
checkBackend();
if (backend == DNN_BACKEND_INFERENCE_ENGINE ||
(backend == DNN_BACKEND_OPENCV && target == DNN_TARGET_OPENCL_FP16))
throw SkipTestException("");
Net net;
{
const string model = findDataFile("dnn/Enet-model-best.net", false);
net = readNetFromTorch(model, true);
ASSERT_TRUE(!net.empty());
}
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
Mat sample = imread(_tf("street.png", false));
Mat inputBlob = blobFromImage(sample, 1./255, Size(), Scalar(), /*swapRB*/true);
net.setInput(inputBlob, "");
Mat out = net.forward();
Mat ref = blobFromNPY(_tf("torch_enet_prob.npy", false));
// Due to numerical instability in Pooling-Unpooling layers (indexes jittering)
// thresholds for ENet must be changed. Accuracy of results was checked on
// Cityscapes dataset and difference in mIOU with Torch is 10E-4%
normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.552);
normAssertSegmentation(ref, out);
const int N = 3;
for (int i = 0; i < N; i++)
{
net.setInput(inputBlob, "");
Mat out = net.forward();
normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.552);
normAssertSegmentation(ref, out);
}
}
TEST_P(Test_ONNX_nets, Googlenet)
{
if (backend == DNN_BACKEND_INFERENCE_ENGINE)
throw SkipTestException("");
const String model = _tf("models/googlenet.onnx");
Net net = readNetFromONNX(model);
ASSERT_FALSE(net.empty());
net.setPreferableBackend(backend);
net.setPreferableTarget(target);
std::vector<Mat> images;
images.push_back( imread(_tf("../googlenet_0.png")) );
images.push_back( imread(_tf("../googlenet_1.png")) );
Mat inp = blobFromImages(images, 1.0f, Size(), Scalar(), false);
Mat ref = blobFromNPY(_tf("../googlenet_prob.npy"));
checkBackend(&inp, &ref);
net.setInput(inp);
ASSERT_FALSE(net.empty());
Mat out = net.forward();
normAssert(ref, out, "", default_l1, default_lInf);
expectNoFallbacksFromIE(net);
}
TEST_P(Reproducibility_MobileNet_SSD, Accuracy)
{
const string proto = findDataFile("dnn/MobileNetSSD_deploy.prototxt", false);
const string model = findDataFile("dnn/MobileNetSSD_deploy.caffemodel", false);
Net net = readNetFromCaffe(proto, model);
int targetId = GetParam();
const float l1 = (targetId == DNN_TARGET_OPENCL_FP16) ? 1.5e-4 : 1e-5;
const float lInf = (targetId == DNN_TARGET_OPENCL_FP16) ? 4e-4 : 1e-4;
net.setPreferableTarget(targetId);
Mat sample = imread(_tf("street.png"));
Mat inp = blobFromImage(sample, 1.0f / 127.5, Size(300, 300), Scalar(127.5, 127.5, 127.5), false);
net.setInput(inp);
Mat out = net.forward();
const float scores_diff = (targetId == DNN_TARGET_OPENCL_FP16) ? 4e-4 : 1e-5;
const float boxes_iou_diff = (targetId == DNN_TARGET_OPENCL_FP16) ? 5e-3 : 1e-4;
Mat ref = blobFromNPY(_tf("mobilenet_ssd_caffe_out.npy"));
normAssertDetections(ref, out, "", 0.0, scores_diff, boxes_iou_diff);
// Check that detections aren't preserved.
inp.setTo(0.0f);
net.setInput(inp);
out = net.forward();
out = out.reshape(1, out.total() / 7);
const int numDetections = out.rows;
ASSERT_NE(numDetections, 0);
for (int i = 0; i < numDetections; ++i)
{
float confidence = out.ptr<float>(i)[2];
ASSERT_EQ(confidence, 0);
}
// Check batching mode.
ref = ref.reshape(1, numDetections);
inp = blobFromImages(std::vector<Mat>(2, sample), 1.0f / 127.5, Size(300, 300), Scalar(127.5, 127.5, 127.5), false);
net.setInput(inp);
Mat outBatch = net.forward();
// Output blob has a shape 1x1x2Nx7 where N is a number of detection for
// a single sample in batch. The first numbers of detection vectors are batch id.
outBatch = outBatch.reshape(1, outBatch.total() / 7);
EXPECT_EQ(outBatch.rows, 2 * numDetections);
normAssert(outBatch.rowRange(0, numDetections), ref, "", l1, lInf);
normAssert(outBatch.rowRange(numDetections, 2 * numDetections).colRange(1, 7), ref.colRange(1, 7),
"", l1, lInf);
}
