mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-25 19:50:38 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

add normalize_bbox layer ocl implementation

Browse Source 
Signed-off-by: Li Peng <peng.li@intel.com>
This commit is contained in:
Li Peng 2018-01-04 19:47:35 +08:00
parent f99a135eda
commit 67f9406cbe
2 changed files with 68 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

63
modules/dnn/src/layers/normalize_bbox_layer.cpp
View File

@ -69,11 +69,74 @@ public:
return true;
}
#ifdef HAVE_OPENCL
bool forward_ocl(InputArrayOfArrays inputs_, OutputArrayOfArrays outputs_, OutputArrayOfArrays internals_)
{
std::vector<UMat> inputs;
std::vector<UMat> outputs;
std::vector<UMat> internals;
inputs_.getUMatVector(inputs);
outputs_.getUMatVector(outputs);
internals_.getUMatVector(internals);
CV_Assert(inputs.size() == 1 && outputs.size() == 1);
CV_Assert(inputs[0].total() == outputs[0].total());
const UMat& inp0 = inputs[0];
UMat& buffer = internals[0];
size_t num = inp0.size[0];
size_t channels = inp0.size[1];
size_t channelSize = inp0.total() / (num * channels);
for (size_t i = 0; i < num; ++i)
{
MatShape s = shape(channels, channelSize);
UMat src = inputs[i].reshape(1, s.size(), &s[0]);
UMat dst = outputs[i].reshape(1, s.size(), &s[0]);
UMat abs_mat;
absdiff(src, cv::Scalar::all(0), abs_mat);
pow(abs_mat, pnorm, buffer);
if (acrossSpatial)
{
// add eps to avoid overflow
float absSum = sum(buffer)[0] + epsilon;
float norm = pow(absSum, 1.0f / pnorm);
multiply(src, 1.0f / norm, dst);
}
if (!blobs.empty())
{
// scale the output
Mat scale = blobs[0];
if (scale.total() == 1)
{
// _scale: 1 x 1
multiply(dst, scale.at<float>(0, 0), dst);
}
else
{
// _scale: _channels x 1
CV_Assert(scale.total() == channels);
repeat(scale, 1, dst.cols, buffer);
multiply(dst, buffer, dst);
}
}
}
return true;
}
#endif
void forward(InputArrayOfArrays inputs_arr, OutputArrayOfArrays outputs_arr, OutputArrayOfArrays internals_arr)
{
CV_TRACE_FUNCTION();
CV_TRACE_ARG_VALUE(name, "name", name.c_str());
CV_OCL_RUN((preferableTarget == DNN_TARGET_OPENCL) &&
OCL_PERFORMANCE_CHECK(ocl::Device::getDefault().isIntel()),
forward_ocl(inputs_arr, outputs_arr, internals_arr))
Layer::forward_fallback(inputs_arr, outputs_arr, internals_arr);
}

5
modules/dnn/test/test_torch_importer.cpp
View File

@ -225,6 +225,11 @@ TEST(Torch_Importer, net_normalize)
runTorchNet("net_normalize", DNN_TARGET_CPU, "", false, true);
}
OCL_TEST(Torch_Importer, net_normalize)
{
runTorchNet("net_normalize", DNN_TARGET_OPENCL, "", false, true);
}
TEST(Torch_Importer, net_padding)
{
runTorchNet("net_padding", DNN_TARGET_CPU, "", false, true);