[ONNX] Enable transposed convolution (#2051)

* [ONNX] Enable transposed convolution

* Review fix pt. 1

* Review fix pt. 2

* Remove unused new_shape

src/ngraph/frontend/onnx_import/CMakeLists.txt

@@ -61,6 +61,8 @@ add_library(onnx_import STATIC
         op/constant.hpp
         op/conv.cpp
         op/conv.hpp
+        op/conv_transpose.cpp
+        op/conv_transpose.hpp
         op/div.hpp
         op/elu.cpp
         op/elu.hpp

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

+//*****************************************************************************
+// Copyright 2017-2018 Intel Corporation
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+//*****************************************************************************
+
+#include <cstdint>
+#include <iterator>
+#include <memory>
+#include <vector>
+
+#include "ngraph/coordinate_diff.hpp"
+#include "ngraph/frontend/onnx_import/op/conv_transpose.hpp"
+#include "ngraph/frontend/onnx_import/utils/broadcasting.hpp"
+#include "ngraph/frontend/onnx_import/utils/convpool.hpp"
+#include "ngraph/op/add.hpp"
+#include "ngraph/op/broadcast.hpp"
+#include "ngraph/op/convolution.hpp"
+#include "ngraph/op/divide.hpp"
+#include "ngraph/op/multiply.hpp"
+#include "ngraph/op/subtract.hpp"
+#include "ngraph/shape.hpp"
+#include "ngraph/strides.hpp"
+
+namespace ngraph
+{
+    namespace onnx_import
+    {
+        namespace op
+        {
+            namespace set_1
+            {
+                NodeVector conv_transpose(const Node& node)
+                {
+                    const NodeVector& inputs = node.get_ng_inputs();
+                    auto data = inputs.at(0);
+                    auto filters = inputs.at(1);
+
+                    const Shape& data_shape = data->get_shape();
+                    const Shape& weights_shape = filters->get_shape();
+                    int num_spatial_dims = data_shape.size() - 2;
+
+                    auto strides = convpool::get_strides(node);
+                    auto dilations = convpool::get_dilations(node);
+                    auto paddings = convpool::get_pads(node);
+                    ngraph::CoordinateDiff padding_below = paddings.first;
+                    ngraph::CoordinateDiff padding_above = paddings.second;
+
+                    Strides data_dilation_strides(num_spatial_dims, 1);
+                    std::vector<std::int64_t> output_shape{
+                        node.get_attribute_value<std::vector<std::int64_t>>("output_shape", {})};
+
+                    std::vector<std::int64_t> output_padding{
+                        node.get_attribute_value<std::vector<std::int64_t>>(
+                            "output_padding", std::vector<std::int64_t>(num_spatial_dims, 0))};
+
+                    Shape data_batch_shape(data_shape.size(), 1);
+                    data_batch_shape[0] = data_shape[0];
+                    data_batch_shape[1] = weights_shape[1];
+
+                    if (!output_shape.empty())
+                    {
+                        if (output_shape.size() > num_spatial_dims)
+                        {
+                            output_shape.erase(std::begin(output_shape),
+                                               std::begin(output_shape) + 2);
+                        }
+                        for (int i = 0; i < num_spatial_dims; ++i)
+                        {
+                            padding_below[i] = (strides[i] * (data_shape[i + 2] - 1) +
+                                                dilations[i] * (weights_shape[i + 2] - 1) -
+                                                data_dilation_strides[i] *
+                                                    (output_shape[i] - output_padding[i] - 1)) /
+                                               2;
+                            if (static_cast<int>(padding_below[i]) < 0)
+                            {
+                                output_padding[i] = -static_cast<int>(padding_below[i]);
+                                padding_below[i] = 0;
+                            }
+                            padding_above[i] = padding_below[i];
+                            data_batch_shape[i + 2] = output_shape[i];
+                        }
+                    }
+                    else
+                    {
+                        for (int i = 0; i < num_spatial_dims && output_shape.empty(); ++i)
+                        {
+                            // Calculating spatial dims of data output shape for ngraph conv backprop op
+                            // | s(ds-1) + d(ws-1) - pb - pa |
+                            // | --------------------------- | + 1 + op
+                            // | _           dds           _ |
+                            //
+                            // d - dilation
+                            // ds - data shape
+                            // dds - data dilation strides
+                            // op - output padding
+                            // pa - padding above
+                            // pb - padding below
+                            // s - strides
+                            // ws - weights shape
+                            data_batch_shape[i + 2] = (strides[i] * (data_shape[i + 2] - 1) +
+                                                       dilations[i] * (weights_shape[i + 2] - 1) -
+                                                       padding_below[i] - padding_above[i]) /
+                                                          data_dilation_strides[i] +
+                                                      1 + output_padding[i];
+                        }
+                    }
+
+                    auto conv_node = std::make_shared<ngraph::op::ConvolutionBackpropData>(
+                        data_batch_shape,
+                        filters,
+                        data,
+                        strides,
+                        dilations,
+                        padding_below,
+                        padding_above,
+                        data_dilation_strides);
+
+                    // no bias param
+                    if (inputs.size() < 3)
+                    {
+                        return {conv_node};
+                    }
+
+                    auto bias = inputs.at(2);
+
+                    return {std::make_shared<ngraph::op::Add>(
+                        conv_node, make_broadcast_node(bias, conv_node->get_shape(), 1))};
+                }
+
+            } // namespace set_1
+
+        } //namespace op
+
+    } // namespace onnx_import
+
+} // namespace ngraph

src/ngraph/frontend/onnx_import/op/conv_transpose.hpp

+//*****************************************************************************
+// Copyright 2017-2018 Intel Corporation
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+//*****************************************************************************
+
+#pragma once
+
+#include "ngraph/node_vector.hpp"
+
+#include "core/node.hpp"
+
+namespace ngraph
+{
+    namespace onnx_import
+    {
+        namespace op
+        {
+            namespace set_1
+            {
+                /// \brief Performs ONNX Transposed Convolution operation.
+                ///
+                /// \param node   The ONNX node object representing this operation.
+                ///
+                /// \return The vector containing Ngraph nodes producing output of ONNX convolution
+                ///         operation.
+                NodeVector conv_transpose(const Node& node);
+
+            } // namespace set_1
+
+        } //namespace op
+
+    } // namespace onnx_import
+
+} // namespace ngraph

src/ngraph/frontend/onnx_import/ops_bridge.cpp

@@ -37,6 +37,7 @@
 #include "op/concat.hpp"
 #include "op/constant.hpp"
 #include "op/conv.hpp"
+#include "op/conv_transpose.hpp"
 #include "op/cos.hpp"
 #include "op/div.hpp"
 #include "op/elu.hpp"
@@ -160,6 +161,7 @@ namespace ngraph
             REGISTER_OPERATOR("Concat", 1, concat);
             REGISTER_OPERATOR("Constant", 1, constant);
             REGISTER_OPERATOR("Conv", 1, conv);
+            REGISTER_OPERATOR("ConvTranspose", 1, conv_transpose);
             REGISTER_OPERATOR("Cos", 1, cos);
             REGISTER_OPERATOR("Div", 1, div);
             REGISTER_OPERATOR("Div", 7, div);