Merge pull request #22017 from xiong-jie-y:py_onnx

Add python bindings for G-API onnx

modules/gapi/CMakeLists.txt

@@ -178,6 +178,7 @@ set(gapi_srcs
 
     # Python bridge
     src/backends/ie/bindings_ie.cpp
+    src/backends/onnx/bindings_onnx.cpp
     src/backends/python/gpythonbackend.cpp
 
     # OpenVPL Streaming source

modules/gapi/include/opencv2/gapi/infer/bindings_onnx.hpp

@@ -0,0 +1,43 @@
+// This file is part of OpenCV project.
+// It is subject to the license terms in the LICENSE file found in the top-level
+// directory of this distribution and at http://opencv.org/license.html.
+
+#ifndef OPENCV_GAPI_INFER_BINDINGS_ONNX_HPP
+#define OPENCV_GAPI_INFER_BINDINGS_ONNX_HPP
+
+#include <opencv2/gapi/gkernel.hpp>     // GKernelPackage
+#include <opencv2/gapi/infer/onnx.hpp>  // Params
+#include "opencv2/gapi/own/exports.hpp"  // GAPI_EXPORTS
+#include <opencv2/gapi/util/any.hpp>
+
+#include <string>
+
+namespace cv {
+namespace gapi {
+namespace onnx {
+
+// NB: Used by python wrapper
+// This class can be marked as SIMPLE, because it's implemented as pimpl
+class GAPI_EXPORTS_W_SIMPLE PyParams {
+public:
+    GAPI_WRAP
+    PyParams() = default;
+
+    GAPI_WRAP
+    PyParams(const std::string& tag, const std::string& model_path);
+
+    GBackend backend() const;
+    std::string tag() const;
+    cv::util::any params() const;
+
+private:
+    std::shared_ptr<Params<cv::gapi::Generic>> m_priv;
+};
+
+GAPI_EXPORTS_W PyParams params(const std::string& tag, const std::string& model_path);
+
+}  // namespace onnx
+}  // namespace gapi
+}  // namespace cv
+
+#endif  // OPENCV_GAPI_INFER_BINDINGS_ONNX_HPP

modules/gapi/include/opencv2/gapi/infer/onnx.hpp

@@ -17,6 +17,7 @@
 
 #include <opencv2/core/cvdef.h>     // GAPI_EXPORTS
 #include <opencv2/gapi/gkernel.hpp> // GKernelPackage
+#include <opencv2/gapi/infer.hpp>   // Generic
 
 namespace cv {
 namespace gapi {
@@ -67,6 +68,8 @@ struct ParamDesc {
     std::vector<bool> normalize; //!< Vector of bool values that enabled or disabled normalize of input data.
 
     std::vector<std::string> names_to_remap; //!< Names of output layers that will be processed in PostProc function.
+
+    bool is_generic;
 };
 } // namespace detail
 
@@ -103,6 +106,7 @@ public:
         desc.model_path = model;
         desc.num_in  = std::tuple_size<typename Net::InArgs>::value;
         desc.num_out = std::tuple_size<typename Net::OutArgs>::value;
+        desc.is_generic = false;
     };
 
     /** @brief Specifies sequence of network input layers names for inference.
@@ -277,6 +281,35 @@ protected:
     detail::ParamDesc desc;
 };
 
+/*
+* @brief This structure provides functions for generic network type that
+* fill inference parameters.
+* @see struct Generic
+*/
+template<>
+class Params<cv::gapi::Generic> {
+public:
+    /** @brief Class constructor.
+
+    Constructs Params based on input information and sets default values for other
+    inference description parameters.
+
+    @param tag string tag of the network for which these parameters are intended.
+    @param model_path path to model file (.onnx file).
+    */
+    Params(const std::string& tag, const std::string& model_path)
+         : desc{model_path, 0u, 0u, {}, {}, {}, {}, {}, {}, {}, {}, {}, true}, m_tag(tag) {}
+
+    // BEGIN(G-API's network parametrization API)
+    GBackend      backend() const { return cv::gapi::onnx::backend(); }
+    std::string   tag()     const { return m_tag; }
+    cv::util::any params()  const { return { desc }; }
+    // END(G-API's network parametrization API)
+protected:
+    detail::ParamDesc desc;
+    std::string m_tag;
+};
+
 } // namespace onnx
 } // namespace gapi
 } // namespace cv

modules/gapi/misc/python/pyopencv_gapi.hpp

@@ -14,6 +14,7 @@
 using gapi_GKernelPackage           = cv::GKernelPackage;
 using gapi_GNetPackage              = cv::gapi::GNetPackage;
 using gapi_ie_PyParams              = cv::gapi::ie::PyParams;
+using gapi_onnx_PyParams            = cv::gapi::onnx::PyParams;
 using gapi_wip_IStreamSource_Ptr    = cv::Ptr<cv::gapi::wip::IStreamSource>;
 using detail_ExtractArgsCallback    = cv::detail::ExtractArgsCallback;
 using detail_ExtractMetaCallback    = cv::detail::ExtractMetaCallback;

modules/gapi/misc/python/shadow_gapi.hpp

@@ -79,5 +79,6 @@ namespace streaming
 namespace detail
 {
     gapi::GNetParam GAPI_EXPORTS_W strip(gapi::ie::PyParams params);
+    gapi::GNetParam GAPI_EXPORTS_W strip(gapi::onnx::PyParams params);
 } // namespace detail
 } // namespace cv

modules/gapi/misc/python/test/test_gapi_infer_onnx.py

@@ -0,0 +1,74 @@
+#!/usr/bin/env python
+
+import numpy as np
+import cv2 as cv
+import os
+import sys
+import unittest
+
+from tests_common import NewOpenCVTests
+
+
+try:
+
+    if sys.version_info[:2] < (3, 0):
+        raise unittest.SkipTest('Python 2.x is not supported')
+
+    CLASSIFICATION_MODEL_PATH = "onnx_models/vision/classification/squeezenet/model/squeezenet1.0-9.onnx"
+
+    testdata_required = bool(os.environ.get('OPENCV_DNN_TEST_REQUIRE_TESTDATA', False))
+
+    class test_gapi_infer(NewOpenCVTests):
+        def find_dnn_file(self, filename, required=None):
+            if not required:
+                required = testdata_required
+            return self.find_file(filename, [os.environ.get('OPENCV_DNN_TEST_DATA_PATH', os.getcwd()),
+                                            os.environ['OPENCV_TEST_DATA_PATH']],
+                                required=required)
+
+        def test_onnx_classification(self):
+            model_path = self.find_dnn_file(CLASSIFICATION_MODEL_PATH)
+
+            if model_path is None:
+                raise unittest.SkipTest("Missing DNN test file")
+
+            in_mat = cv.imread(
+                self.find_file("cv/dpm/cat.png",
+                [os.environ.get('OPENCV_TEST_DATA_PATH')]))
+
+            g_in = cv.GMat()
+            g_infer_inputs = cv.GInferInputs()
+            g_infer_inputs.setInput("data_0", g_in)
+            g_infer_out = cv.gapi.infer("squeeze-net", g_infer_inputs)
+            g_out = g_infer_out.at("softmaxout_1")
+
+            comp = cv.GComputation(cv.GIn(g_in), cv.GOut(g_out))
+
+            net = cv.gapi.onnx.params("squeeze-net", model_path)
+            try:
+                out_gapi = comp.apply(cv.gin(in_mat), cv.gapi.compile_args(cv.gapi.networks(net)))
+            except cv.error as err:
+                if err.args[0] == "G-API has been compiled without ONNX support":
+                    raise unittest.SkipTest("G-API has been compiled without ONNX support")
+                else:
+                    raise
+
+            self.assertEqual((1, 1000, 1, 1), out_gapi.shape)
+
+
+except unittest.SkipTest as e:
+
+    message = str(e)
+
+    class TestSkip(unittest.TestCase):
+        def setUp(self):
+            self.skipTest('Skip tests: ' + message)
+
+        def test_skip():
+            pass
+
+    pass
+
+
+if __name__ == '__main__':
+    NewOpenCVTests.bootstrap()

modules/gapi/src/backends/onnx/bindings_onnx.cpp

@@ -0,0 +1,24 @@
+// This file is part of OpenCV project.
+// It is subject to the license terms in the LICENSE file found in the top-level
+// directory of this distribution and at http://opencv.org/license.html.
+
+#include <opencv2/gapi/infer/bindings_onnx.hpp>
+
+cv::gapi::onnx::PyParams::PyParams(const std::string& tag,
+                                   const std::string& model_path)
+    : m_priv(std::make_shared<Params<cv::gapi::Generic>>(tag, model_path)) {}
+
+cv::gapi::GBackend cv::gapi::onnx::PyParams::backend() const {
+    return m_priv->backend();
+}
+
+std::string cv::gapi::onnx::PyParams::tag() const { return m_priv->tag(); }
+
+cv::util::any cv::gapi::onnx::PyParams::params() const {
+    return m_priv->params();
+}
+
+cv::gapi::onnx::PyParams cv::gapi::onnx::params(
+    const std::string& tag, const std::string& model_path) {
+    return {tag, model_path};
+}

modules/gapi/src/backends/onnx/gonnxbackend.cpp

@@ -735,7 +735,8 @@ void ONNXCompiled::extractMat(ONNXCallContext &ctx, const size_t in_idx, Views&
     }
 }
 
-void ONNXCompiled::setOutput(int i, cv::Mat &m) {
+void ONNXCompiled::setOutput(int i, cv::Mat &m)
+{
     // FIXME: No need in double-indexing?
     out_data[i] = m;
 }
@@ -1133,9 +1134,34 @@ namespace {
             // FIXME: Introduce a DNNBackend interface which'd specify
             // the framework for this???
             GONNXModel gm(gr);
-            const auto &np = gm.metadata(nh).get<NetworkParams>();
-            const auto &pp = cv::util::any_cast<cv::gapi::onnx::detail::ParamDesc>(np.opaque);
+            auto &np = gm.metadata(nh).get<NetworkParams>();
+            auto &pp = cv::util::any_cast<cv::gapi::onnx::detail::ParamDesc>(np.opaque);
             const auto &ki = cv::util::any_cast<KImpl>(ii.opaque);
+
+            GModel::Graph model(gr);
+            auto& op = model.metadata(nh).get<Op>();
+            if (pp.is_generic) {
+                auto& info = cv::util::any_cast<cv::detail::InOutInfo>(op.params);
+
+                for (const auto& a : info.in_names)
+                {
+                    pp.input_names.push_back(a);
+                }
+                // Adding const input is necessary because the definition of input_names
+                // includes const input.
+                for (const auto& a : pp.const_inputs)
+                {
+                    pp.input_names.push_back(a.first);
+                }
+                pp.num_in = info.in_names.size();
+
+                for (const auto& a : info.out_names)
+                {
+                    pp.output_names.push_back(a);
+                }
+                pp.num_out = info.out_names.size();
+            }
+
             gm.metadata(nh).set(ONNXUnit{pp});
             gm.metadata(nh).set(ONNXCallable{ki.run});
             gm.metadata(nh).set(CustomMetaFunction{ki.customMetaFunc});