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

Merge pull request #24378 from fengyuentau:instance_norm

Browse Source 
dnn onnx: add instance norm layer #24378

Resolves https://github.com/opencv/opencv/issues/24377
Relates https://github.com/opencv/opencv/pull/24092#discussion_r1349841644

| Perf | multi-thread | single-thread |
| - | - | - |
| x: [2, 64, 180, 240] | 3.95ms | 11.12ms |

Todo:

- [x] speed up by multi-threading
- [x] add perf
- [x] add backend: OpenVINO
- [x] add backend: CUDA
- [x] add backend: OpenCL (no fp16)
- [ ] add backend: CANN (will be done via https://github.com/opencv/opencv/pull/24462)


### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake

```
force_builders=Linux OpenCL,Win64 OpenCL,Custom
buildworker:Custom=linux-4
build_image:Custom=ubuntu:18.04
modules_filter:Custom=none
disable_ipp:Custom=ON
```
This commit is contained in:
Yuantao Feng
2023-11-07 17:59:10 +08:00
committed by GitHub
parent 832f738db0
commit ee0822dc4d
10 changed files with 449 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/perf/perf_layer.cpp
+57
View File
@@ -683,6 +683,62 @@ PERF_TEST_P_(Layer_GatherElements, GatherElements)
test_layer({2700, 1, 2914}, {2700, 1, 81}, 2);
}
struct Layer_InstanceNorm : public TestBaseWithParam<tuple<Backend, Target> >
{
void test_layer(const std::vector<int>& x_shape)
{
int backendId = get<0>(GetParam());
int targetId = get<1>(GetParam());
Mat x(x_shape, CV_32FC1);
Mat scale(x_shape[1], 1, CV_32FC1);
Mat b(x_shape[1], 1, CV_32FC1);
randu(x, 0.f, 1.f);
randu(scale, 0.f, 1.f);
randu(b, 0.f, 1.f);
Net net;
LayerParams lp;
lp.type = "InstanceNormalization";
lp.name = "testLayer";
int id = net.addLayerToPrev(lp.name, lp.type, lp);
net.connect(0, 0, id, 0);
net.connect(0, 1, id, 1);
net.connect(0, 2, id, 2);
// warmup
{
std::vector<String> inpNames{"x", "scale", "b"};
net.setInputsNames(inpNames);
net.setInput(x, inpNames[0]);
net.setInput(scale, inpNames[1]);
net.setInput(b, inpNames[2]);
net.setPreferableBackend(backendId);
net.setPreferableTarget(targetId);
Mat out = net.forward();
}
TEST_CYCLE()
{
Mat res = net.forward();
}
SANITY_CHECK_NOTHING();
}
int N = 2;
int C = 64;
int H = 180;
int W = 240;
};
PERF_TEST_P_(Layer_InstanceNorm, InstanceNorm)
{
test_layer({N, C, H, W});
}
INSTANTIATE_TEST_CASE_P(/**/, Layer_Slice, dnnBackendsAndTargets(false, false));
INSTANTIATE_TEST_CASE_P(/**/, Layer_NaryEltwise, testing::Values(std::make_tuple(DNN_BACKEND_OPENCV, DNN_TARGET_CPU)));
#ifdef HAVE_CUDA
@@ -693,6 +749,7 @@ INSTANTIATE_TEST_CASE_P(/**/, Layer_ScatterND, testing::Values(std::make_tuple(D
INSTANTIATE_TEST_CASE_P(/**/, Layer_LayerNorm, testing::Values(std::make_tuple(DNN_BACKEND_OPENCV, DNN_TARGET_CPU)));
INSTANTIATE_TEST_CASE_P(/**/, Layer_LayerNormExpanded, testing::Values(std::make_tuple(DNN_BACKEND_OPENCV, DNN_TARGET_CPU)));
INSTANTIATE_TEST_CASE_P(/**/, Layer_GatherElements, testing::Values(std::make_tuple(DNN_BACKEND_OPENCV, DNN_TARGET_CPU)));
INSTANTIATE_TEST_CASE_P(/**/, Layer_InstanceNorm, testing::Values(std::make_tuple(DNN_BACKEND_OPENCV, DNN_TARGET_CPU)));
typedef TestBaseWithParam<tuple<Vec4i, int, bool, tuple<Backend, Target> > > Layer_FullyConnected;