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); } }
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(Torch_Importer, ENet_accuracy) { Net net; { const string model = findDataFile("dnn/Enet-model-best.net", false); net = readNetFromTorch(model, true); ASSERT_FALSE(net.empty()); } 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(Reproducibility_FCN, Accuracy) { Net net; { const string proto = findDataFile("dnn/fcn8s-heavy-pascal.prototxt", false); const string model = findDataFile("dnn/fcn8s-heavy-pascal.caffemodel", false); net = readNetFromCaffe(proto, model); ASSERT_FALSE(net.empty()); } Mat sample = imread(_tf("street.png")); ASSERT_TRUE(!sample.empty()); std::vector<int> layerIds; std::vector<size_t> weights, blobs; net.getMemoryConsumption(shape(1,3,227,227), layerIds, weights, blobs); net.setInput(blobFromImage(sample, 1.0f, Size(500, 500), Scalar(), false), "data"); Mat out = net.forward("score"); Mat refData = imread(_tf("caffe_fcn8s_prob.png"), IMREAD_ANYDEPTH); int shape[] = {1, 21, 500, 500}; Mat ref(4, shape, CV_32FC1, refData.data); normAssert(ref, out); }
TEST(readNet, Regression) { Net net = readNet(findDataFile("dnn/squeezenet_v1.1.prototxt", false), findDataFile("dnn/squeezenet_v1.1.caffemodel", false)); EXPECT_FALSE(net.empty()); net = readNet(findDataFile("dnn/opencv_face_detector.caffemodel", false), findDataFile("dnn/opencv_face_detector.prototxt", false)); EXPECT_FALSE(net.empty()); net = readNet(findDataFile("dnn/openface_nn4.small2.v1.t7", false)); EXPECT_FALSE(net.empty()); net = readNet(findDataFile("dnn/tiny-yolo-voc.cfg", false), findDataFile("dnn/tiny-yolo-voc.weights", false)); EXPECT_FALSE(net.empty()); net = readNet(findDataFile("dnn/ssd_mobilenet_v1_coco.pbtxt", false), findDataFile("dnn/ssd_mobilenet_v1_coco.pb", false)); EXPECT_FALSE(net.empty()); }
TEST(Reproducibility_YoloVoc, Accuracy) { Net net; { const string cfg = findDataFile("dnn/yolo-voc.cfg", false); const string model = findDataFile("dnn/yolo-voc.weights", false); net = readNetFromDarknet(cfg, model); ASSERT_FALSE(net.empty()); } // 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/yolo-voc.cfg ./yolo-voc.weights -thresh 0.24 ./dog416.png // There are 3 objects (6-car, 1-bicycle, 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[] = { 0.740161F, 0.214100F, 0.325575F, 0.173418F, 0.750769F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.750469F, 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, 0.501618F, 0.504757F, 0.461713F, 0.481310F, 0.783550F, 0.000000F, 0.780879F, 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, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.279968F, 0.638651F, 0.282737F, 0.600284F, 0.901864F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.901615F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F, 0.000000F }; const int number_of_objects = 3; Mat ref(number_of_objects, sizeof(ref_array) / (number_of_objects * sizeof(float)), CV_32FC1, &ref_array); normAssert(ref, detection); }
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); }
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(Test_Caffe, memory_read) { const string proto = findDataFile("dnn/bvlc_googlenet.prototxt", false); const string model = findDataFile("dnn/bvlc_googlenet.caffemodel", false); string dataProto; ASSERT_TRUE(readFileInMemory(proto, dataProto)); string dataModel; ASSERT_TRUE(readFileInMemory(model, dataModel)); Net net = readNetFromCaffe(dataProto.c_str(), dataProto.size()); ASSERT_FALSE(net.empty()); Net net2 = readNetFromCaffe(dataProto.c_str(), dataProto.size(), dataModel.c_str(), dataModel.size()); ASSERT_FALSE(net2.empty()); }
TEST(Reproducibility_AlexNet, Accuracy) { Net net; { const string proto = findDataFile("dnn/bvlc_alexnet.prototxt", false); const string model = findDataFile("dnn/bvlc_alexnet.caffemodel", false); net = readNetFromCaffe(proto, model); ASSERT_FALSE(net.empty()); } Mat sample = imread(_tf("grace_hopper_227.png")); ASSERT_TRUE(!sample.empty()); net.setInput(blobFromImage(sample, 1.0f, Size(227, 227), Scalar(), false), "data"); Mat out = net.forward("prob"); Mat ref = blobFromNPY(_tf("caffe_alexnet_prob.npy")); normAssert(ref, out); }
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(Test_TensorFlow, inception_accuracy) { Net net; { const string model = findDataFile("dnn/tensorflow_inception_graph.pb", false); net = readNetFromTensorflow(model); ASSERT_FALSE(net.empty()); } net.setPreferableBackend(DNN_BACKEND_OPENCV); Mat sample = imread(_tf("grace_hopper_227.png")); ASSERT_TRUE(!sample.empty()); Mat inputBlob = blobFromImage(sample, 1.0, Size(224, 224), Scalar(), /*swapRB*/true); net.setInput(inputBlob, "input"); Mat out = net.forward("softmax2"); Mat ref = blobFromNPY(_tf("tf_inception_prob.npy")); normAssert(ref, out); }
TEST(Test_TensorFlow, inception_accuracy) { Net net; { const string model = findDataFile("dnn/tensorflow_inception_graph.pb", false); net = readNetFromTensorflow(model); ASSERT_FALSE(net.empty()); } Mat sample = imread(_tf("grace_hopper_227.png")); ASSERT_TRUE(!sample.empty()); resize(sample, sample, Size(224, 224)); Mat inputBlob = blobFromImage(sample); net.setInput(inputBlob, "input"); Mat out = net.forward("softmax2"); Mat ref = blobFromNPY(_tf("tf_inception_prob.npy")); normAssert(ref, out); }
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); }
TEST(Test_TensorFlow, read_inception) { Net net; { const string model = findDataFile("dnn/tensorflow_inception_graph.pb", false); net = readNetFromTensorflow(model); ASSERT_FALSE(net.empty()); } Mat sample = imread(_tf("grace_hopper_227.png")); ASSERT_TRUE(!sample.empty()); Mat input; resize(sample, input, Size(224, 224)); input -= 128; // mean sub Mat inputBlob = blobFromImage(input); net.setInput(inputBlob, "input"); Mat out = net.forward("softmax2"); std::cout << out.dims << std::endl; }
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_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); } }
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(Reproducibility_SSD, Accuracy) { Net net; { const string proto = findDataFile("dnn/ssd_vgg16.prototxt", false); const string model = findDataFile("dnn/VGG_ILSVRC2016_SSD_300x300_iter_440000.caffemodel", false); net = readNetFromCaffe(proto, model); ASSERT_FALSE(net.empty()); } Mat sample = imread(_tf("street.png")); ASSERT_TRUE(!sample.empty()); if (sample.channels() == 4) cvtColor(sample, sample, COLOR_BGRA2BGR); Mat in_blob = blobFromImage(sample, 1.0f, Size(300, 300), Scalar(), false); net.setInput(in_blob, "data"); Mat out = net.forward("detection_out"); Mat ref = blobFromNPY(_tf("ssd_out.npy")); normAssert(ref, out); }
TEST_P(Reproducibility_AlexNet, Accuracy) { bool readFromMemory = get<0>(GetParam()); Net net; { const string proto = findDataFile("dnn/bvlc_alexnet.prototxt", false); const string model = findDataFile("dnn/bvlc_alexnet.caffemodel", false); if (readFromMemory) { string dataProto; ASSERT_TRUE(readFileInMemory(proto, dataProto)); string dataModel; ASSERT_TRUE(readFileInMemory(model, dataModel)); net = readNetFromCaffe(dataProto.c_str(), dataProto.size(), dataModel.c_str(), dataModel.size()); } else net = readNetFromCaffe(proto, model); ASSERT_FALSE(net.empty()); } int targetId = get<1>(GetParam()); const float l1 = 1e-5; const float lInf = (targetId == DNN_TARGET_OPENCL_FP16) ? 3e-3 : 1e-4; net.setPreferableTarget(targetId); Mat sample = imread(_tf("grace_hopper_227.png")); ASSERT_TRUE(!sample.empty()); net.setInput(blobFromImage(sample, 1.0f, Size(227, 227), Scalar(), false), "data"); Mat out = net.forward("prob"); Mat ref = blobFromNPY(_tf("caffe_alexnet_prob.npy")); normAssert(ref, out, "", l1, lInf); }
TEST(Test_Darknet, read_yolo_voc) { Net net = readNetFromDarknet(_tf("yolo-voc.cfg")); ASSERT_FALSE(net.empty()); }
TEST(Test_Caffe, read_gtsrb) { Net net = readNetFromCaffe(_tf("gtsrb.prototxt")); ASSERT_FALSE(net.empty()); }
TEST(Test_Caffe, read_googlenet) { Net net = readNetFromCaffe(_tf("bvlc_googlenet.prototxt")); ASSERT_FALSE(net.empty()); }
TEST(Torch_Importer, simple_read) { Net net; ASSERT_NO_THROW(net = readNetFromTorch(_tf("net_simple_net.txt"), false)); ASSERT_FALSE(net.empty()); }
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); }