potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-30 15:53:03 +04:00
Code Packages Releases Wiki Activity

fp16 ocl support for googlenet

Browse Source 
Signed-off-by: Li Peng <peng.li@intel.com>
This commit is contained in:
Li Peng
2018-04-26 19:22:29 +08:00
parent 329abb5b64
commit 3dd916882a
32 changed files with 2480 additions and 303 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/src/layers/elementwise_layers.cpp
+73 -8
View File
@@ -176,7 +176,7 @@ public:
{
CV_TRACE_FUNCTION();
CV_OCL_RUN((this->preferableTarget == DNN_TARGET_OPENCL) &&
CV_OCL_RUN(IS_DNN_OPENCL_TARGET(this->preferableTarget) &&
OCL_PERFORMANCE_CHECK(ocl::Device::getDefault().isIntel()),
func.applyOCL(inputs_arr, outputs_arr, internals_arr))
@@ -223,7 +223,12 @@ public:
#ifdef HAVE_OPENCL
static String oclGetTMacro(const UMat &m)
{
return String("-DT=") + ocl::typeToStr(m.type()) + String(" ");
String str_name = ocl::typeToStr(m.type());
if (str_name == "short")
str_name = "half";
return format("-DT=%s -Dconvert_T=convert_%s ", str_name.c_str(), str_name.c_str());
}
#endif
@@ -516,8 +521,28 @@ struct SigmoidFunctor
#ifdef HAVE_OPENCL
bool applyOCL(InputArrayOfArrays inps, OutputArrayOfArrays outs, OutputArrayOfArrays internals)
{
// TODO: implement OCL version
return false;
std::vector<UMat> inputs;
std::vector<UMat> outputs;
inps.getUMatVector(inputs);
outs.getUMatVector(outputs);
String buildopt = oclGetTMacro(inputs[0]);
for (size_t i = 0; i < inputs.size(); i++)
{
UMat& src = inputs[i];
UMat& dst = outputs[i];
ocl::Kernel kernel("SigmoidForward", ocl::dnn::activations_oclsrc, buildopt);
kernel.set(0, (int)src.total());
kernel.set(1, ocl::KernelArg::PtrReadOnly(src));
kernel.set(2, ocl::KernelArg::PtrWriteOnly(dst));
size_t gSize = src.total();
CV_Assert(kernel.run(1, &gSize, NULL, false));
}
return true;
}
#endif
@@ -561,8 +586,28 @@ struct ELUFunctor
#ifdef HAVE_OPENCL
bool applyOCL(InputArrayOfArrays inps, OutputArrayOfArrays outs, OutputArrayOfArrays internals)
{
// TODO: implement OCL version
return false;
std::vector<UMat> inputs;
std::vector<UMat> outputs;
inps.getUMatVector(inputs);
outs.getUMatVector(outputs);
String buildopt = oclGetTMacro(inputs[0]);
for (size_t i = 0; i < inputs.size(); i++)
{
UMat& src = inputs[i];
UMat& dst = outputs[i];
ocl::Kernel kernel("ELUForward", ocl::dnn::activations_oclsrc, buildopt);
kernel.set(0, (int)src.total());
kernel.set(1, ocl::KernelArg::PtrReadOnly(src));
kernel.set(2, ocl::KernelArg::PtrWriteOnly(dst));
size_t gSize = src.total();
CV_Assert(kernel.run(1, &gSize, NULL, false));
}
return true;
}
#endif
@@ -604,8 +649,28 @@ struct AbsValFunctor
#ifdef HAVE_OPENCL
bool applyOCL(InputArrayOfArrays inps, OutputArrayOfArrays outs, OutputArrayOfArrays internals)
{
// TODO: implement OCL version
return false;
std::vector<UMat> inputs;
std::vector<UMat> outputs;
inps.getUMatVector(inputs);
outs.getUMatVector(outputs);
String buildopt = oclGetTMacro(inputs[0]);
for (size_t i = 0; i < inputs.size(); i++)
{
UMat& src = inputs[i];
UMat& dst = outputs[i];
ocl::Kernel kernel("AbsValForward", ocl::dnn::activations_oclsrc, buildopt);
kernel.set(0, (int)src.total());
kernel.set(1, ocl::KernelArg::PtrReadOnly(src));
kernel.set(2, ocl::KernelArg::PtrWriteOnly(dst));
size_t gSize = src.total();
CV_Assert(kernel.run(1, &gSize, NULL, false));
}
return true;
}
#endif