Add support for input dynamic shape in ONNX LpNorm operator (#4613)

ngraph/frontend/onnx_import/src/op/lp_norm.cpp

@@ -42,8 +42,6 @@ namespace ngraph
                     const auto data_shape = data.get_partial_shape();
                     const auto data_rank = data_shape.rank();
 
-                    CHECK_VALID_NODE(
-                        node, data_shape.is_static(), "Data shape must be static for lp_norm op");
                     const auto data_rank_value = data_rank.get_length();
                     const std::int64_t p_norm{node.get_attribute_value<std::int64_t>("p", 2)};
 
@@ -62,8 +60,7 @@ namespace ngraph
                     std::shared_ptr<ngraph::Node> norm = ngraph::builder::opset1::lp_norm(
                         data, normalize_axis_const, static_cast<std::size_t>(p_norm));
 
-                    const auto target_shape = default_opset::Constant::create(
-                        element::i64, Shape{size_t(data_rank_value)}, data_shape.to_shape());
+                    const auto target_shape = std::make_shared<default_opset::ShapeOf>(data);
 
                     // Create a default axes order matching the data tensor rank and erase the
                     // element at the 'normalize_axis' position. The erased element indicates the

ngraph/test/models/onnx/lp_norm_default_dynamic.prototxt

@@ -0,0 +1,50 @@
+ir_version: 3
+producer_name: "nGraph ONNX Importer"
+graph {
+  node {
+    input: "x"
+    output: "y"
+    op_type: "LpNormalization"
+  }
+  name: "lp_norm_graph"
+  input {
+    name: "x"
+    type {
+      tensor_type {
+        elem_type: 1
+        shape {
+          dim {
+          }
+          dim {
+            dim_value: 3
+          }
+          dim {
+            dim_value: 4
+          }
+        }
+      }
+    }
+  }
+  output {
+    name: "y"
+    type {
+      tensor_type {
+        elem_type: 1
+        shape {
+          dim {
+            dim_value: 2
+          }
+          dim {
+            dim_value: 3
+          }
+          dim {
+            dim_value: 4
+          }
+        }
+      }
+    }
+  }
+}
+opset_import {
+  version: 1
+}

ngraph/test/onnx/onnx_import.in.cpp

@@ -2886,6 +2886,26 @@ NGRAPH_TEST(${BACKEND_NAME}, onnx_model_lp_norm_default)
     test_case.run();
 }
 
+NGRAPH_TEST(${BACKEND_NAME}, onnx_model_lp_norm_default_dynamic)
+{
+    const auto function = onnx_import::import_onnx_model(
+        file_util::path_join(SERIALIZED_ZOO, "onnx/lp_norm_default_dynamic.prototxt"));
+
+    Shape data_shape{2, 3, 4};
+    std::vector<float> data(shape_size(data_shape));
+    std::iota(std::begin(data), std::end(data), 1);
+
+    auto test_case = test::TestCase<TestEngine, test::TestCaseType::DYNAMIC>(function);
+    test_case.add_input<float>(data_shape, data);
+    test_case.add_expected_output<float>(
+        data_shape, {0.18257418f, 0.36514837f, 0.5477225f,  0.73029673f, 0.37904903f, 0.45485884f,
+                     0.5306686f,  0.60647845f, 0.42616236f, 0.47351375f, 0.5208651f,  0.5682165f,
+                     0.4469492f,  0.48132992f, 0.51571065f, 0.5500913f,  0.45862272f, 0.48560053f,
+                     0.5125783f,  0.53955615f, 0.46609157f, 0.4882864f,  0.51048124f, 0.5326761f});
+
+    test_case.run();
+}
+
 NGRAPH_TEST(${BACKEND_NAME}, onnx_model_instance_normalization)
 {
     const auto function = onnx_import::import_onnx_model(

ngraph/test/runtime/ie/unit_test.manifest

@@ -202,6 +202,7 @@ onnx_model_range_positive_step
 onnx_model_range_negative_step
 onnx_dyn_shapes_slice_1_3d_input_21_axes_ends_max
 onnx_model_max_pool_dyn_rank_without_default_attrs
+onnx_model_lp_norm_default_dynamic
 
 # (Constant W, R inputs are required) Ticket: 49207
 # Function references undeclared parameters