Merge pull request opencv#16983 from dkurt:dnn_tf_prelu

Author: alalek

modules/dnn/src/tensorflow/tf_graph_simplifier.cpp

@@ -223,6 +223,26 @@ class FlattenShapeSubgraph : public Subgraph
     }
 };
 
+class FlattenProdSubgraph : public Subgraph
+{
+public:
+    FlattenProdSubgraph()
+    {
+        int input = addNodeToMatch("");
+        int shape = addNodeToMatch("Shape", input);
+        int stack = addNodeToMatch("Const");
+        int stack_1 = addNodeToMatch("Const");
+        int stack_2 = addNodeToMatch("Const");
+        int strided_slice = addNodeToMatch("StridedSlice", shape, stack, stack_1, stack_2);
+        int prod = addNodeToMatch("Prod", strided_slice, addNodeToMatch("Const"));
+        int shape_pack = addNodeToMatch("Const");
+        int pack = addNodeToMatch("Pack", shape_pack, prod);
+        addNodeToMatch("Reshape", input, pack);
+
+        setFusedNode("Flatten", input);
+    }
+};
+
 // K.layers.Softmax
 class SoftMaxKerasSubgraph : public Subgraph
 {
@@ -629,6 +649,36 @@ class KerasMVNSubgraph : public TFSubgraph
     }
 };
 
+class PReLUSubgraph : public TFSubgraph
+{
+public:
+    PReLUSubgraph(bool negativeScales_) : negativeScales(negativeScales_)
+    {
+        int input = addNodeToMatch("");
+        int scales = addNodeToMatch("Const");
+        int neg = addNodeToMatch("Neg", input);
+        int relu_neg = addNodeToMatch("Relu", neg);
+        int finalScales = negativeScales ? addNodeToMatch("Neg", scales) : scales;
+        int mul = addNodeToMatch("Mul", finalScales, relu_neg);
+        int relu_pos = addNodeToMatch("Relu", input);
+        addNodeToMatch("Add", relu_pos, mul);
+        setFusedNode("PReLU", input, scales);
+    }
+
+    virtual void finalize(tensorflow::GraphDef&, tensorflow::NodeDef* fusedNode,
+                          std::vector<tensorflow::NodeDef*>& inputNodes) CV_OVERRIDE
+    {
+        if (!negativeScales)
+        {
+            Mat scales = getTensorContent(inputNodes[1]->attr().at("value").tensor(), /*copy*/false);
+            scales *= -1;
+        }
+    }
+
+private:
+    bool negativeScales;
+};
+
 void simplifySubgraphs(tensorflow::GraphDef& net)
 {
     std::vector<Ptr<Subgraph> > subgraphs;
@@ -649,6 +699,16 @@ void simplifySubgraphs(tensorflow::GraphDef& net)
     subgraphs.push_back(Ptr<Subgraph>(new SoftMaxSlimV2Subgraph()));
     subgraphs.push_back(Ptr<Subgraph>(new ReshapeAsShapeSubgraph()));
     subgraphs.push_back(Ptr<Subgraph>(new KerasMVNSubgraph()));
+    subgraphs.push_back(Ptr<Subgraph>(new PReLUSubgraph(true)));
+    subgraphs.push_back(Ptr<Subgraph>(new PReLUSubgraph(false)));
+    subgraphs.push_back(Ptr<Subgraph>(new FlattenProdSubgraph()));
+
+    for (int i = 0; i < net.node_size(); ++i)
+    {
+        tensorflow::NodeDef* layer = net.mutable_node(i);
+        if (layer->op() == "AddV2")
+            layer->set_op("Add");
+    }
 
     simplifySubgraphs(Ptr<ImportGraphWrapper>(new TFGraphWrapper(net)), subgraphs);
 }

modules/dnn/src/tensorflow/tf_importer.cpp

@@ -1231,6 +1231,7 @@ void TFImporter::populateNet(Net dstNet)
                 // Only NHWC <-> NCHW permutations are allowed. OpenCV is always
                 // keep NCHW layout this way.
                 int inpLayout = getDataLayout(layer.input(0), data_layouts);
+                std::string type = "Identity";
                 if (inpLayout == DATA_LAYOUT_NHWC)
                 {
                     if (permData[0] == 0 && permData[1] == 3 && permData[2] == 1 && permData[3] == 2)
@@ -1245,6 +1246,15 @@ void TFImporter::populateNet(Net dstNet)
                         // in OpenCV: NCHW->NCHW
                         data_layouts[name] = DATA_LAYOUT_NHWC;
                     }
+                    else if (permData[0] == 0 && permData[1] == 3 && permData[2] == 2 && permData[3] == 1)
+                    {
+                        // in TensorFlow: NHWC->NCWH
+                        // in OpenCV: NCHW->NCWH
+                        int permData[] = {0, 1, 3, 2};
+                        layerParams.set("order", DictValue::arrayInt<int*>(permData, perm.total()));
+                        data_layouts[name] = DATA_LAYOUT_NCHW;  // we keep track NCHW because channels position only matters
+                        type = "Permute";
+                    }
                     else
                         CV_Error(Error::StsParseError, "Only NHWC <-> NCHW permutations are allowed.");
                 }
@@ -1265,7 +1275,7 @@ void TFImporter::populateNet(Net dstNet)
                     else
                         CV_Error(Error::StsParseError, "Only NHWC <-> NCHW permutations are allowed.");
                 }
-                int id = dstNet.addLayer(name, "Identity", layerParams);
+                int id = dstNet.addLayer(name, type, layerParams);
                 layer_id[name] = id;
                 connect(layer_id, dstNet, parsePin(layer.input(0)), id, 0);
             }

modules/dnn/test/test_tf_importer.cpp

@@ -956,11 +956,25 @@ TEST_P(Test_TensorFlow_layers, resize_bilinear)
     runTensorFlowNet("resize_bilinear_factor");
 }
 
-TEST_P(Test_TensorFlow_layers, tf2_keras)
+TEST_P(Test_TensorFlow_layers, tf2_dense)
 {
     runTensorFlowNet("tf2_dense");
 }
 
+TEST_P(Test_TensorFlow_layers, tf2_prelu)
+{
+    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);
+    runTensorFlowNet("tf2_prelu");
+}
+
+TEST_P(Test_TensorFlow_layers, tf2_permute_nhwc_ncwh)
+{
+    runTensorFlowNet("tf2_permute_nhwc_ncwh");
+}
+
 TEST_P(Test_TensorFlow_layers, squeeze)
 {
 #if defined(INF_ENGINE_RELEASE)