Merge pull request #13002 from dkurt:onnx_padding

modules/dnn/src/onnx/onnx_importer.cpp

@@ -96,6 +96,8 @@ Mat getMatFromTensor(opencv_onnx::TensorProto& tensor_proto)
     for (int i = 0; i < tensor_proto.dims_size(); i++) {
             sizes.push_back(tensor_proto.dims(i));
     }
+    if (sizes.empty())
+        sizes.assign(1, 1);
     if (datatype == opencv_onnx::TensorProto_DataType_FLOAT) {
 
         if (!tensor_proto.float_data().empty()) {
@@ -173,11 +175,31 @@ LayerParams ONNXImporter::getLayerParams(const opencv_onnx::NodeProto& node_prot
         }
         else if(attribute_name == "pads")
         {
-            CV_Assert(attribute_proto.ints_size() == 4);
-            lp.set("pad_t", saturate_cast<int32_t>(attribute_proto.ints(0)));
-            lp.set("pad_l", saturate_cast<int32_t>(attribute_proto.ints(1)));
-            lp.set("pad_b", saturate_cast<int32_t>(attribute_proto.ints(2)));
-            lp.set("pad_r", saturate_cast<int32_t>(attribute_proto.ints(3)));
+            if (node_proto.op_type() == "Pad")
+            {
+                // Padding layer.
+                // Paddings are in order begin0, begin1, .. beginN, end0, end1, ..., endN.
+                // We need to shuffle it to begin0, end0, begin1, end1, ...
+                CV_Assert(attribute_proto.ints_size() % 2 == 0);
+                const int dims = attribute_proto.ints_size() / 2;
+                std::vector<int32_t> paddings;
+                paddings.reserve(attribute_proto.ints_size());
+                for (int i = 0; i < dims; ++i)
+                {
+                    paddings.push_back(attribute_proto.ints(i));
+                    paddings.push_back(attribute_proto.ints(dims + i));
+                }
+                lp.set("paddings", DictValue::arrayInt(&paddings[0], paddings.size()));
+            }
+            else
+            {
+                // Convolution or pooling.
+                CV_Assert(attribute_proto.ints_size() == 4);
+                lp.set("pad_t", saturate_cast<int32_t>(attribute_proto.ints(0)));
+                lp.set("pad_l", saturate_cast<int32_t>(attribute_proto.ints(1)));
+                lp.set("pad_b", saturate_cast<int32_t>(attribute_proto.ints(2)));
+                lp.set("pad_r", saturate_cast<int32_t>(attribute_proto.ints(3)));
+            }
         }
         else if(attribute_name == "auto_pad")
         {
@@ -543,6 +565,10 @@ void ONNXImporter::populateNet(Net dstNet)
                 replaceLayerParam(layerParams, "shape", "dim");
             }
         }
+        else if (layer_type == "Pad")
+        {
+            layerParams.type = "Padding";
+        }
         else
         {
             for (int j = 0; j < node_proto.input_size(); j++) {

modules/dnn/test/test_onnx_importer.cpp

@@ -129,6 +129,11 @@ TEST_P(Test_ONNX_layers, Constant)
     testONNXModels("constant");
 }
 
+TEST_P(Test_ONNX_layers, Padding)
+{
+    testONNXModels("padding");
+}
+
 TEST_P(Test_ONNX_layers, MultyInputs)
 {
     const String model =  _tf("models/multy_inputs.onnx");