potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-31 08:13:04 +04:00
Code Packages Releases Wiki Activity

Merge pull request #15537 from l-bat:ngraph

Browse Source 
* Support nGraph

* Fix resize
This commit is contained in:
Lubov Batanina
2019-12-02 16:16:06 +03:00
committed by Alexander Alekhin
parent 20ac7f40f6
commit 7523c777c5
49 changed files with 3189 additions and 447 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/src/layers/scale_layer.cpp
+31 -1
View File
@@ -13,6 +13,8 @@ Implementation of Scale layer.
#include "layers_common.hpp"
#include "../op_halide.hpp"
#include "../op_inf_engine.hpp"
#include "../ie_ngraph.hpp"
#include <opencv2/dnn/shape_utils.hpp>
namespace cv
@@ -51,7 +53,7 @@ public:
virtual bool supportBackend(int backendId) CV_OVERRIDE
{
return backendId == DNN_BACKEND_OPENCV || backendId == DNN_BACKEND_HALIDE ||
(backendId == DNN_BACKEND_INFERENCE_ENGINE && axis == 1);
((backendId == DNN_BACKEND_INFERENCE_ENGINE_NN_BUILDER_2019 || backendId == DNN_BACKEND_INFERENCE_ENGINE_NGRAPH) && axis == 1);
}
void forward(InputArrayOfArrays inputs_arr, OutputArrayOfArrays outputs_arr, OutputArrayOfArrays internals_arr) CV_OVERRIDE
@@ -222,6 +224,34 @@ public:
}
#endif // HAVE_INF_ENGINE
#ifdef HAVE_DNN_NGRAPH
virtual Ptr<BackendNode> initNgraph(const std::vector<Ptr<BackendWrapper> >& inputs, const std::vector<Ptr<BackendNode> >& nodes) CV_OVERRIDE
{
CV_Assert(!blobs.empty());
const size_t numChannels = blobs[0].total();
auto ieInpNode = nodes[0].dynamicCast<InfEngineNgraphNode>()->node;
std::vector<size_t> shape(ieInpNode->get_shape().size(), 1);
shape[1] = numChannels;
auto weight = hasWeights ?
std::make_shared<ngraph::op::Constant>(ngraph::element::f32,
ngraph::Shape(shape), blobs[0].data) :
std::make_shared<ngraph::op::Constant>(ngraph::element::f32,
ngraph::Shape(shape), std::vector<float>(numChannels, 1).data());
auto bias = hasBias ?
std::make_shared<ngraph::op::Constant>(ngraph::element::f32,
ngraph::Shape(shape), blobs.back().data) :
std::make_shared<ngraph::op::Constant>(ngraph::element::f32,
ngraph::Shape(shape), std::vector<float>(numChannels, 0).data());
auto scale_node = std::make_shared<ngraph::op::v1::Multiply>(ieInpNode, weight, ngraph::op::AutoBroadcastType::NUMPY);
auto scale_shift = std::make_shared<ngraph::op::v1::Add>(scale_node, bias, ngraph::op::AutoBroadcastType::NUMPY);
return Ptr<BackendNode>(new InfEngineNgraphNode(scale_shift));
}
#endif // HAVE_DNN_NGRAPH
void getScaleShift(Mat& scale, Mat& shift) const CV_OVERRIDE
{
scale = hasWeights ? blobs[0] : Mat();