mirror of
https://github.com/open-mmlab/mmdeploy.git
synced 2025-01-14 08:09:43 +08:00
36124f6205
* check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
59 lines
1.9 KiB
C++
59 lines
1.9 KiB
C++
// Copyright (c) OpenMMLab. All rights reserved.
|
|
|
|
#include <cuda_runtime.h>
|
|
|
|
#include "preprocess/transform/image2tensor.h"
|
|
#include "preprocess/transform/transform_utils.h"
|
|
|
|
namespace mmdeploy {
|
|
namespace cuda {
|
|
|
|
template <typename T>
|
|
void Transpose(const T* src, int height, int width, int channels, T* dst, cudaStream_t stream);
|
|
|
|
class ImageToTensorImpl final : public ::mmdeploy::ImageToTensorImpl {
|
|
public:
|
|
explicit ImageToTensorImpl(const Value& args) : ::mmdeploy::ImageToTensorImpl(args) {}
|
|
|
|
protected:
|
|
Result<Tensor> HWC2CHW(const Tensor& tensor) override {
|
|
OUTCOME_TRY(auto src_tensor, MakeAvailableOnDevice(tensor, device_, stream_));
|
|
auto h = tensor.shape(1);
|
|
auto w = tensor.shape(2);
|
|
auto c = tensor.shape(3);
|
|
auto hw = h * w;
|
|
|
|
Tensor dst_tensor(src_tensor.desc());
|
|
dst_tensor.Reshape({1, c, h, w});
|
|
|
|
auto stream = ::mmdeploy::GetNative<cudaStream_t>(stream_);
|
|
|
|
if (DataType::kINT8 == tensor.data_type()) {
|
|
auto input = src_tensor.data<uint8_t>();
|
|
auto output = dst_tensor.data<uint8_t>();
|
|
Transpose(input, (int)h, (int)w, (int)c, output, stream);
|
|
} else if (DataType::kFLOAT == tensor.data_type()) {
|
|
auto input = src_tensor.data<float>();
|
|
auto output = dst_tensor.data<float>();
|
|
Transpose(input, (int)h, (int)w, (int)c, output, stream);
|
|
} else {
|
|
assert(0);
|
|
}
|
|
return dst_tensor;
|
|
}
|
|
};
|
|
|
|
class ImageToTensorImplCreator : public Creator<::mmdeploy::ImageToTensorImpl> {
|
|
public:
|
|
const char* GetName() const override { return "cuda"; }
|
|
int GetVersion() const override { return 1; }
|
|
ReturnType Create(const Value& cfg) override { return std::make_unique<ImageToTensorImpl>(cfg); }
|
|
};
|
|
|
|
} // namespace cuda
|
|
} // namespace mmdeploy
|
|
|
|
using ::mmdeploy::ImageToTensorImpl;
|
|
using ::mmdeploy::cuda::ImageToTensorImplCreator;
|
|
REGISTER_MODULE(ImageToTensorImpl, ImageToTensorImplCreator);
|