mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-12-17 19:08:01 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity
e64c5dc4c6
opencv/modules/dnn/src/init.cpp

240 lines
12 KiB
C++
Raw Normal View History

dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2013, OpenCV Foundation, all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include "precomp.hpp"
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
#include <opencv2/dnn/layer.details.hpp>
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
Merge pull request #23604 from dkurt:dnn_no_protobuf Build DNN without Protobuf DNN module can be built without Protobuf for Darknet, TFLite, OpenVINO, Torch (not PyTorch) models. ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_PROTOBUF=OFF \ -DWITH_OPENCL=OFF 7.1M lib/libopencv_dnn.so.4.7.0 ``` ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_OPENCL=OFF 3.9M lib/libopencv_dnn.so.4.7.0 ``` ### 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
2023-05-15 17:23:18 +08:00
#if defined(HAVE_PROTOBUF) && !defined(BUILD_PLUGIN)
dnn: protobuf shutdown
2017-07-25 21:13:56 +08:00
#include <google/protobuf/stubs/common.h>
dnn: plugin support for OpenVINO
2022-03-18 14:19:30 +08:00
#endif
dnn: protobuf shutdown
2017-07-25 21:13:56 +08:00
dnn: added "hidden" experimental namespace Main purpose of this namespace is to avoid using of incompatible binaries that will cause applications crashes. This additional namespace will not impact "Source code API". This change allows to maintain ABI checks (with easy filtering out).
2017-06-29 03:59:02 +08:00
namespace cv {
namespace dnn {
dnn: update "guard" inline namespace - differ from 3.4 branch
2018-09-04 04:20:02 +08:00
CV__DNN_INLINE_NS_BEGIN
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
static Mutex* __initialization_mutex = NULL;
Mutex& getInitializationMutex()
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
{
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
if (__initialization_mutex == NULL)
__initialization_mutex = new Mutex();
return *__initialization_mutex;
}
// force initialization (single-threaded environment)
Mutex* __initialization_mutex_initializer = &getInitializationMutex();
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
Merge pull request #23604 from dkurt:dnn_no_protobuf Build DNN without Protobuf DNN module can be built without Protobuf for Darknet, TFLite, OpenVINO, Torch (not PyTorch) models. ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_PROTOBUF=OFF \ -DWITH_OPENCL=OFF 7.1M lib/libopencv_dnn.so.4.7.0 ``` ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_OPENCL=OFF 3.9M lib/libopencv_dnn.so.4.7.0 ``` ### 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
2023-05-15 17:23:18 +08:00
#if defined(HAVE_PROTOBUF) && !defined(BUILD_PLUGIN)
dnn: protobuf shutdown
2017-07-25 21:13:56 +08:00
namespace {
using namespace google::protobuf;
class ProtobufShutdown {
public:
bool initialized;
ProtobufShutdown() : initialized(true) {}
~ProtobufShutdown()
{
initialized = false;
google::protobuf::ShutdownProtobufLibrary();
}
};
} // namespace
dnn: plugin support for OpenVINO
2022-03-18 14:19:30 +08:00
#endif
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
void initializeLayerFactory()
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
{
dnn: added trace macros
2017-06-28 19:46:58 +08:00
CV_TRACE_FUNCTION();
Merge pull request #23604 from dkurt:dnn_no_protobuf Build DNN without Protobuf DNN module can be built without Protobuf for Darknet, TFLite, OpenVINO, Torch (not PyTorch) models. ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_PROTOBUF=OFF \ -DWITH_OPENCL=OFF 7.1M lib/libopencv_dnn.so.4.7.0 ``` ``` cmake \ -DCMAKE_BUILD_TYPE=Release \ -DBUILD_LIST=dnn \ -DWITH_OPENCL=OFF 3.9M lib/libopencv_dnn.so.4.7.0 ``` ### 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
2023-05-15 17:23:18 +08:00
#if defined(HAVE_PROTOBUF) && !defined(BUILD_PLUGIN)
Utilize CV_UNUSED macro
2018-09-07 19:33:52 +08:00
static ProtobufShutdown protobufShutdown; CV_UNUSED(protobufShutdown);
dnn: plugin support for OpenVINO
2022-03-18 14:19:30 +08:00
#endif
dnn: protobuf shutdown
2017-07-25 21:13:56 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Slice, SliceLayer);
CV_DNN_REGISTER_LAYER_CLASS(Split, SplitLayer);
CV_DNN_REGISTER_LAYER_CLASS(Concat, ConcatLayer);
CV_DNN_REGISTER_LAYER_CLASS(Reshape, ReshapeLayer);
CV_DNN_REGISTER_LAYER_CLASS(Flatten, FlattenLayer);
Added ResizeBilinear op for tf (#11050) * Added ResizeBilinear op for tf Combined ResizeNearestNeighbor and ResizeBilinear layers into Resize (with an interpolation param). Minor changes to tf_importer and resize layer to save some code lines Minor changes in init.cpp Minor changes in tf_importer.cpp * Replaced implementation of a custom ResizeBilinear layer to all layers * Use Mat::ptr. Replace interpolation flags
2018-06-07 21:29:04 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Resize, ResizeLayer);
Implement Interp layer using Resize layer
2018-06-23 00:26:47 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Interp, InterpLayer);
Faster-RCNN object detection models from TensorFlow
2018-04-03 23:28:05 +08:00
CV_DNN_REGISTER_LAYER_CLASS(CropAndResize, CropAndResizeLayer);
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Convolution, ConvolutionLayer);
CV_DNN_REGISTER_LAYER_CLASS(Deconvolution, DeconvolutionLayer);
CV_DNN_REGISTER_LAYER_CLASS(Pooling, PoolingLayer);
ROIPooling layer
2017-12-04 23:45:30 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ROIPooling, PoolingLayer);
R-FCN models support
2017-12-20 00:43:49 +08:00
CV_DNN_REGISTER_LAYER_CLASS(PSROIPooling, PoolingLayer);
Support for some reduce layers for onnx
2022-03-18 10:19:13 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Reduce, ReduceLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(LRN, LRNLayer);
CV_DNN_REGISTER_LAYER_CLASS(InnerProduct, InnerProductLayer);
dnn: add gemm_layer in place of fully_connected_layer for onnx models (#23897) * first commit * turned C from input to constant; force C constant in impl; better handling 0d/1d cases * integrate with gemm from ficus nn * fix const inputs * adjust threshold for int8 tryQuantize * adjust threshold for int8 quantized 2 * support batched gemm and matmul; tune threshold for rcnn_ilsvrc13; update googlenet * add gemm perf against innerproduct * add perf tests for innerproduct with bias * fix perf * add memset * renamings for next step * add dedicated perf gemm * add innerproduct in perf_gemm * remove gemm and innerproduct perf tests from perf_layer * add perf cases for vit sizes; prepack constants * remove batched gemm; fix wrong trans; optimize KC * remove prepacking for const A; several fixes for const B prepacking * add todos and gemm expression * add optimized branch for avx/avx2 * trigger build * update macros and signature * update signature * fix macro * fix bugs for neon aarch64 & x64 * add backends: cuda, cann, inf_ngraph and vkcom * fix cuda backend * test commit for cuda * test cuda backend * remove debug message from cuda backend * use cpu dispatcher * fix neon macro undef in dispatcher * fix dispatcher * fix inner kernel for neon aarch64 * fix compiling issue on armv7; try fixing accuracy issue on other platforms * broadcast C with beta multiplied; improve func namings * fix bug for avx and avx2 * put all platform-specific kernels in dispatcher * fix typos * attempt to fix compile issues on x64 * run old gemm when neon, avx, avx2 are all not available; add kernel for armv7 neon * fix typo * quick fix: add macros for pack4 * quick fix: use vmlaq_f32 for armv7 * quick fix for missing macro of fast gemm pack f32 4 * disable conformance tests when optimized branches are not supported * disable perf tests when optimized branches are not supported * decouple cv_try_neon and cv_neon_aarch64 * drop googlenet_2023; add fastGemmBatched * fix step in fastGemmBatched * cpu: fix initialization ofb; gpu: support batch * quick followup fix for cuda * add default kernels * quick followup fix to avoid macro redef * optmized kernels for lasx * resolve mis-alignment; remove comments * tune performance for x64 platform * tune performance for neon aarch64 * tune for armv7 * comment time consuming tests * quick follow-up fix
2023-09-20 05:53:34 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Gemm, GemmLayer);
Merge pull request #24694 from fengyuentau:matmul_refactor dnn: refactor ONNX MatMul with fastGemm #24694 Done: - [x] add backends - [x] CUDA - [x] OpenVINO - [x] CANN - [x] OpenCL - [x] Vulkan - [x] add perf tests - [x] const B case ### Benchmark Tests are done on M1. All data is in milliseconds (ms). | Configuration | MatMul (Prepacked) | MatMul | InnerProduct | | - | - | - | - | | A=[12, 197, 197], B=[12, 197, 64], trans_a=0, trans_b=0 | **0.39** | 0.41 | 1.33 | | A=[12, 197, 64], B=[12, 64, 197], trans_a=0, trans_b=0 | **0.42** | 0.42 | 1.17 | | A=[12, 50, 64], B=[12, 64, 50], trans_a=0, trans_b=0 | **0.13** | 0.15 | 0.33 | | A=[12, 50, 50], B=[12, 50, 64], trans_a=0, trans_b=0 | **0.11** | 0.13 | 0.22 | | A=[16, 197, 197], B=[16, 197, 64], trans_a=0, trans_b=0 | **0.46** | 0.54 | 1.46 | | A=[16, 197, 64], B=[16, 64, 197], trans_a=0, trans_b=0 | **0.46** | 0.95 | 1.74 | | A=[16, 50, 64], B=[16, 64, 50], trans_a=0, trans_b=0 | **0.18** | 0.32 | 0.43 | | A=[16, 50, 50], B=[16, 50, 64], trans_a=0, trans_b=0 | **0.15** | 0.25 | 0.25 | ### 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
2023-12-20 00:36:41 +08:00
CV_DNN_REGISTER_LAYER_CLASS(MatMul, MatMulLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Softmax, SoftmaxLayer);
Case sensitive dnn layers types
2020-03-22 23:50:15 +08:00
CV_DNN_REGISTER_LAYER_CLASS(SoftMax, SoftmaxLayer); // For compatibility. See https://github.com/opencv/opencv/issues/16877
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(MVN, MVNLayer);
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ReLU, ReLULayer);
Layers for MobileNet from TensorFlow
2017-09-14 03:18:02 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ReLU6, ReLU6Layer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ChannelsPReLU, ChannelsPReLULayer);
PReLU layer from Caffe
2017-10-09 22:47:46 +08:00
CV_DNN_REGISTER_LAYER_CLASS(PReLU, ChannelsPReLULayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Sigmoid, SigmoidLayer);
CV_DNN_REGISTER_LAYER_CLASS(TanH, TanHLayer);
Merge pull request #15808 from thebhatman:Mish_swish * Added Swish and Mish activations * Fixed whitespace errors * Kernel implementation done * Added function for launching kernel * Changed type of 1.0 * Attempt to add test for Swish and Mish * Resolving type mismatch for log * exp from device * Use log1pexp instead of adding 1 * Added openCL kernels
2019-12-02 05:06:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Swish, SwishLayer);
CV_DNN_REGISTER_LAYER_CLASS(Mish, MishLayer);
Added ELU and test for it
2017-08-01 21:58:34 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ELU, ELULayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(BNLL, BNLLLayer);
CV_DNN_REGISTER_LAYER_CLASS(AbsVal, AbsLayer);
CV_DNN_REGISTER_LAYER_CLASS(Power, PowerLayer);
Added exp layer
2021-02-21 00:46:00 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Exp, ExpLayer);
add Ceil, Floor, Log, Round, Sqrt, Not, Equal, Less, Greater
2021-10-14 22:38:57 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Ceil, CeilLayer);
CV_DNN_REGISTER_LAYER_CLASS(Floor, FloorLayer);
CV_DNN_REGISTER_LAYER_CLASS(Log, LogLayer);
CV_DNN_REGISTER_LAYER_CLASS(Round, RoundLayer);
CV_DNN_REGISTER_LAYER_CLASS(Sqrt, SqrtLayer);
CV_DNN_REGISTER_LAYER_CLASS(Not, NotLayer);
Add acos, acosh, asin, asinh, atan, atanh, cos, cosh, erf, hardswish, sin, sinh, softplus, softsign, tan layers
2021-12-02 20:45:27 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Acos, AcosLayer);
CV_DNN_REGISTER_LAYER_CLASS(Acosh, AcoshLayer);
CV_DNN_REGISTER_LAYER_CLASS(Asin, AsinLayer);
CV_DNN_REGISTER_LAYER_CLASS(Asinh, AsinhLayer);
CV_DNN_REGISTER_LAYER_CLASS(Atan, AtanLayer);
CV_DNN_REGISTER_LAYER_CLASS(Atanh, AtanhLayer);
CV_DNN_REGISTER_LAYER_CLASS(Cos, CosLayer);
CV_DNN_REGISTER_LAYER_CLASS(Cosh, CoshLayer);
CV_DNN_REGISTER_LAYER_CLASS(Erf, ErfLayer);
CV_DNN_REGISTER_LAYER_CLASS(HardSwish, HardSwishLayer);
CV_DNN_REGISTER_LAYER_CLASS(Sin, SinLayer);
CV_DNN_REGISTER_LAYER_CLASS(Sinh, SinhLayer);
add ONNX OP sign, shrink and reciprocal
2022-04-07 15:32:12 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Sign, SignLayer);
CV_DNN_REGISTER_LAYER_CLASS(Shrink, ShrinkLayer);
Add acos, acosh, asin, asinh, atan, atanh, cos, cosh, erf, hardswish, sin, sinh, softplus, softsign, tan layers
2021-12-02 20:45:27 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Softplus, SoftplusLayer);
CV_DNN_REGISTER_LAYER_CLASS(Softsign, SoftsignLayer);
CV_DNN_REGISTER_LAYER_CLASS(Tan, TanLayer);
add celu, hardsigmoid, selu, thresholdedrelu layers
2021-12-04 04:17:07 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Celu, CeluLayer);
CV_DNN_REGISTER_LAYER_CLASS(HardSigmoid, HardSigmoidLayer);
CV_DNN_REGISTER_LAYER_CLASS(Selu, SeluLayer);
CV_DNN_REGISTER_LAYER_CLASS(ThresholdedRelu,ThresholdedReluLayer);
Merge pull request #23219 from fengyuentau:add_gelu Add GELU layer for vision transformers * add gelu and gelu approximation * drop setKernelParams
2023-02-11 02:03:29 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Gelu, GeluLayer);
CV_DNN_REGISTER_LAYER_CLASS(GeluApproximation, GeluApproximationLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(BatchNorm, BatchNormLayer);
CV_DNN_REGISTER_LAYER_CLASS(MaxUnpool, MaxUnpoolLayer);
CV_DNN_REGISTER_LAYER_CLASS(Dropout, BlankLayer);
CV_DNN_REGISTER_LAYER_CLASS(Identity, BlankLayer);
Grayscale colorization model (https://github.com/richzhang/colorization) test.
2017-10-05 18:04:22 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Silence, BlankLayer);
Concat layer from TensorFlow with constant inputs
2018-12-04 20:02:22 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Const, ConstLayer);
add ArgMax and ArgMin layers
2021-12-07 00:33:59 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Arg, ArgLayer);
add ONNX OP sign, shrink and reciprocal
2022-04-07 15:32:12 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Reciprocal, ReciprocalLayer);
add Gather implementation
2022-03-16 23:41:39 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Gather, GatherLayer);
Merge pull request #24092 from Aser-Abdelfatah:GSoC_Support_GatherElements_ONNX GSoC Add ONNX Support for GatherElements #24092 Merge with: https://github.com/opencv/opencv_extra/pull/1082 Adds support to the ONNX operator GatherElements [operator docs](https://github.com/onnx/onnx/blob/main/docs/Operators.md#GatherElements) Added tests to opencv_extra at pull request https://github.com/opencv/opencv_extra/pull/1082 ### 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
2023-10-18 15:41:47 +08:00
CV_DNN_REGISTER_LAYER_CLASS(GatherElements, GatherElementsLayer);
Merge pull request #23047 from fengyuentau:layer_norm dnn: add layer normalization for vision transformers * add layer norm onnx parser, impl and tests * add onnx graph simplifier for layer norm expanded * handle the case when constants are of type Initializer * add test case for layer norm expanded with initializers * use CV_Assert & CV_CheckType in place of CV_Assert_N; use forward_fallback for OCL_FP16 * use const ref / ref in parameters of invoker::run; extract inner const if from nested loop; use size_t in place of ull * template hasBias * remove trailing whitespace * use pointer parameter with null check; move normSize division & mean_square division outside of loop; use std::max to ensure positive value before std::sqrt * refactor implementation, optimize parallel_for * disable layer norm expanded * remove the removal of layer norm optional outputs
2023-01-27 21:35:59 +08:00
CV_DNN_REGISTER_LAYER_CLASS(LayerNormalization, LayerNormLayer);
dnn: expand refactor with cv::broadcast for onnx models (#24295) * add expand impl with cv::broadcast * remove expandMid * deduce shape from -1 * add constant folding * handle input constant; handle input constant 1d * add expand conformance tests; add checks to disallow shape of neg values; add early copy for unchanged total elements * fix ExpandSubgraph * dummy commit to trigger build * dummy commit to trigger build 1 * remove conformance from test names
2023-09-27 14:28:52 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Expand, ExpandLayer);
Merge pull request #24378 from fengyuentau:instance_norm 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 ```
2023-11-07 17:59:10 +08:00
CV_DNN_REGISTER_LAYER_CLASS(InstanceNormalization, InstanceNormLayer);
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Crop, CropLayer);
CV_DNN_REGISTER_LAYER_CLASS(Eltwise, EltwiseLayer);
Merge pull request #21865 from rogday:nary_eltwise_layers Reimplementation of Element-wise layers with broadcasting support * init * semi-working initial version * add small_vector * wip * remove smallvec * add nary function * replace auto with Mat in lambda expr used in transform * uncomment asserts * autobuffer shape_buf & step_buf * fix a missing bracket * fixed a missing addLayer in parseElementWise * solve one-dimensional broadcast * remove pre_broadcast_transform for the case of two constants; fix missing constBlobsExtraInfo when addConstant is called * one autobuffer for step & shape * temporal fix for the missing original dimension information * fix parseUnsqueeze when it gets a 1d tensor constant * support sum/mean/min/max with only one input * reuse old code to handle cases of two non-constant inputs * add condition to handle div & mul of two non-constant inputs * use || instead of or * remove trainling spaces * enlarge buf in binary_forward to contain other buffer * use autobuffer in nary_forward * generate data randomly and add more cases for perf * add op and, or & xor * update perf_dnn * remove some comments * remove legacy; add two ONNX conformance tests in filter * move from cpu_denylist to all_denylist * adjust parsing for inputs>=2 Co-authored-by: fengyuentau <yuantao.feng@opencv.org.cn>
2022-07-19 11:14:05 +08:00
CV_DNN_REGISTER_LAYER_CLASS(NaryEltwise, NaryEltwiseLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Permute, PermuteLayer);
Add ShuffleChannel layer
2018-06-19 19:35:07 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ShuffleChannel, ShuffleChannelLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(PriorBox, PriorBoxLayer);
Update PriorBox layer
2017-11-27 20:59:31 +08:00
CV_DNN_REGISTER_LAYER_CLASS(PriorBoxClustered, PriorBoxLayer);
Added DNN Darknet Yolo v2 for object detection
2017-09-25 04:34:08 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Reorg, ReorgLayer);
CV_DNN_REGISTER_LAYER_CLASS(Region, RegionLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(DetectionOutput, DetectionOutputLayer);
CV_DNN_REGISTER_LAYER_CLASS(NormalizeBBox, NormalizeBBoxLayer);
CV_DNN_REGISTER_LAYER_CLASS(Normalize, NormalizeBBoxLayer);
CV_DNN_REGISTER_LAYER_CLASS(Shift, ShiftLayer);
CV_DNN_REGISTER_LAYER_CLASS(Padding, PaddingLayer);
Faster-RCNN models support
2017-12-14 00:06:30 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Proposal, ProposalLayer);
dnn: fix LayerFactory initialization
2017-06-28 01:34:17 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Scale, ScaleLayer);
add Ceil, Floor, Log, Round, Sqrt, Not, Equal, Less, Greater
2021-10-14 22:38:57 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Compare, CompareLayer);
Merge pull request #16575 from l-bat:flownet2 Support FlowNet2 model * Support DataAugmentation layer * Fix warnings * Fix comments * Support Correlation layer * TEST * Support Correlation layer * Supported Accum and FlowWarp layers * Supported ChannelNorm layer * Supported Resample with inputs.size() > 1 * Fixed comments * Refactoring * Added tests * Add resample test * Added asserts in resize layer * Updated DataAugmentation layer * Update convolution layer * Refactoring * Fix data augmentation layer * Fix caffe importer * Fix resize * Switch to Mat ptr * Remove useless resize type * Used ResizeLayer in Accum * Split ChannelNormLayer * Delete duplicate assert * Add sample * Fix sample * Added colormap
2020-05-19 20:29:50 +08:00
CV_DNN_REGISTER_LAYER_CLASS(DataAugmentation, DataAugmentationLayer);
CV_DNN_REGISTER_LAYER_CLASS(Correlation, CorrelationLayer);
CV_DNN_REGISTER_LAYER_CLASS(Accum, AccumLayer);
CV_DNN_REGISTER_LAYER_CLASS(FlowWarp, FlowWarpLayer);
LSTM layer for TensorFlow importer
2017-08-25 19:45:03 +08:00
CV_DNN_REGISTER_LAYER_CLASS(LSTM, LSTMLayer);
Merge pull request #20442 from JulieBar:gru_layer * Add initialization and inference for GRU layer * fix issues found on review
2021-08-07 15:07:37 +08:00
CV_DNN_REGISTER_LAYER_CLASS(GRU, GRULayer);
Merge pull request #20483 from thezane:support-cumsum-layer-for-onnx * Support cumsum layer for onnx * Add unit tests * Address review comments
2021-08-18 01:09:25 +08:00
CV_DNN_REGISTER_LAYER_CLASS(CumSum, CumSumLayer);
Merge pull request #24037 from Abdurrahheem:ash/dev_einsum Add Support for Einsum Layer #24037 ### This PR adding support for [Einsum Layer](https://pytorch.org/docs/stable/generated/torch.einsum.html) (in progress). This PR is currently not to be merged but only reviewed. Test cases are located in [#1090](https://github.com/opencv/opencv_extra/pull/1090)RP in OpenCV extra **DONE**: - [x] 2-5D GMM support added - [x] Matrix transpose support added - [x] Reduction type comupte 'ij->j' - [x] 2nd shape computation - during forward **Next PRs**: - [ ] Broadcasting reduction "...ii ->...i" - [ ] Add lazy shape deduction. "...ij, ...jk->...ik" - [ ] Add implicit output computation support. "bij,bjk ->" (output subscripts should be "bik") - [ ] Add support for CUDA backend - [ ] BatchWiseMultiply optimize **Later in 5.x version (requires support for 1D matrices)**: - [ ] Add 1D vector multiplication support - [ ] Inter product "i, i" (problems with 1D shapes) ### 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 - [ ] 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
2023-09-22 16:25:02 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Einsum, EinsumLayer);
int8 layers and 8-bit quantization support
2021-08-19 12:26:47 +08:00
implementation of scatter and scatternd with conformance tests enabled
2022-09-18 22:13:55 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Scatter, ScatterLayer);
CV_DNN_REGISTER_LAYER_CLASS(ScatterND, ScatterNDLayer);
tile impl
2022-11-15 14:29:30 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Tile, TileLayer);
implementation of scatter and scatternd with conformance tests enabled
2022-09-18 22:13:55 +08:00
int8 layers and 8-bit quantization support
2021-08-19 12:26:47 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Quantize, QuantizeLayer);
CV_DNN_REGISTER_LAYER_CLASS(Dequantize, DequantizeLayer);
Merge pull request #20535 from SamFC10:onnx-q dnn : int8 quantized layers support in onnx importer * added quantized layers support in onnx importer * added more cases in eltwise node, some more checks * added tests for quantized nodes * relax thresholds for failed tests, address review comments * refactoring based on review comments * added support for unsupported cases and pre-quantized resnet50 test * relax thresholds due to int8 resize layer
2021-10-05 02:07:38 +08:00
CV_DNN_REGISTER_LAYER_CLASS(Requantize, RequantizeLayer);
int8 layers and 8-bit quantization support
2021-08-19 12:26:47 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ConvolutionInt8, ConvolutionLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(InnerProductInt8, InnerProductLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(PoolingInt8, PoolingLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(EltwiseInt8, EltwiseLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(BatchNormInt8, BatchNormLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ScaleInt8, ScaleLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ShiftInt8, ShiftLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ReLUInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ReLU6Int8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(SigmoidInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(TanHInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(SwishInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(MishInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ELUInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(BNLLInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(AbsValInt8, ActivationLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(SoftmaxInt8, SoftmaxLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(SoftMaxInt8, SoftmaxLayerInt8);
CV_DNN_REGISTER_LAYER_CLASS(ConcatInt8, ConcatLayer);
CV_DNN_REGISTER_LAYER_CLASS(FlattenInt8, FlattenLayer);
CV_DNN_REGISTER_LAYER_CLASS(PaddingInt8, PaddingLayer);
CV_DNN_REGISTER_LAYER_CLASS(BlankInt8, BlankLayer);
CV_DNN_REGISTER_LAYER_CLASS(DropoutInt8, BlankLayer);
CV_DNN_REGISTER_LAYER_CLASS(IdentityInt8, BlankLayer);
CV_DNN_REGISTER_LAYER_CLASS(SilenceInt8, BlankLayer);
CV_DNN_REGISTER_LAYER_CLASS(ConstInt8, ConstLayer);
CV_DNN_REGISTER_LAYER_CLASS(ReshapeInt8, ReshapeLayer);
Merge pull request #20535 from SamFC10:onnx-q dnn : int8 quantized layers support in onnx importer * added quantized layers support in onnx importer * added more cases in eltwise node, some more checks * added tests for quantized nodes * relax thresholds for failed tests, address review comments * refactoring based on review comments * added support for unsupported cases and pre-quantized resnet50 test * relax thresholds due to int8 resize layer
2021-10-05 02:07:38 +08:00
CV_DNN_REGISTER_LAYER_CLASS(ResizeInt8, ResizeLayer);
int8 layers and 8-bit quantization support
2021-08-19 12:26:47 +08:00
CV_DNN_REGISTER_LAYER_CLASS(SplitInt8, SplitLayer);
CV_DNN_REGISTER_LAYER_CLASS(SliceInt8, SliceLayer);
CV_DNN_REGISTER_LAYER_CLASS(CropInt8, CropLayer);
CV_DNN_REGISTER_LAYER_CLASS(PermuteInt8, PermuteLayer);
CV_DNN_REGISTER_LAYER_CLASS(ReorgInt8, ReorgLayer);
CV_DNN_REGISTER_LAYER_CLASS(ShuffleChannelInt8, ShuffleChannelLayer);
dnn: move module from opencv_contrib https://github.com/opencv/opencv_contrib/tree/e6f63c7a38ca40c5dc33e38736e3027e3528d6cb/modules/dnn
2017-06-26 18:35:51 +08:00
}
dnn: update "guard" inline namespace - differ from 3.4 branch
2018-09-04 04:20:02 +08:00
CV__DNN_INLINE_NS_END
dnn: added "hidden" experimental namespace Main purpose of this namespace is to avoid using of incompatible binaries that will cause applications crashes. This additional namespace will not impact "Source code API". This change allows to maintain ABI checks (with easy filtering out).
2017-06-29 03:59:02 +08:00
}} // namespace
Reference in New Issue Copy Permalink