[Doc] Add C++ comments for external models (PaddlePaddle#394)

* first commit for yolov7 * pybind for yolov7 * CPP README.md * CPP README.md * modified yolov7.cc * README.md * python file modify * delete license in fastdeploy/ * repush the conflict part * README.md modified * README.md modified * file path modified * file path modified * file path modified * file path modified * file path modified * README modified * README modified * move some helpers to private * add examples for yolov7 * api.md modified * api.md modified * api.md modified * YOLOv7 * yolov7 release link * yolov7 release link * yolov7 release link * copyright * change some helpers to private * change variables to const and fix documents. * gitignore * Transfer some funtions to private member of class * Transfer some funtions to private member of class * Merge from develop (#9) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <[email protected]> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> * first commit for yolor * for merge * Develop (#11) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <[email protected]> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> * Yolor (#16) * Develop (#11) (#12) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <[email protected]> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> * Develop (#13) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <[email protected]> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * Develop (#14) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <[email protected]> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> Co-authored-by: Jason <[email protected]> * add is_dynamic for YOLO series (#22) * modify ppmatting backend and docs * modify ppmatting docs * fix the PPMatting size problem * fix LimitShort's log * retrigger ci * modify PPMatting docs * modify the way for dealing with LimitShort * add C++ comments for external models * add comments for external models * add comments for ocr, seg, ppyoloe and yolov5cls * modify copyright to pass the code style check Co-authored-by: Jason <[email protected]> Co-authored-by: root <[email protected]> Co-authored-by: DefTruth <[email protected]> Co-authored-by: huangjianhui <[email protected]> Co-authored-by: Jason <[email protected]>
ziqi-jin · Oct 21, 2022 · 640dfcf · 640dfcf
1 parent 542d42a
commit 640dfcf
Show file tree

Hide file tree

Showing 29 changed files with 566 additions and 203 deletions.
diff --git a/fastdeploy/vision/classification/contrib/yolov5cls.h b/fastdeploy/vision/classification/contrib/yolov5cls.h
@@ -44,7 +44,7 @@ class FASTDEPLOY_DECL YOLOv5Cls : public FastDeployModel {
 
   /** \brief Predict the classification result for an input image
    *
-   * \param[in] im The input image data, comes from cv::imread()
+   * \param[in] im The input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
    * \param[in] result The output classification result will be writen to this structure
    * \param[in] topk Returns the topk classification result with the highest predicted probability, the default is 1
    * \return true if the prediction successed, otherwise false

diff --git a/fastdeploy/vision/detection/contrib/nanodet_plus.h b/fastdeploy/vision/detection/contrib/nanodet_plus.h
@@ -23,34 +23,51 @@ namespace fastdeploy {
 namespace vision {
 
 namespace detection {
-
+/*! @brief NanoDetPlus model object used when to load a NanoDetPlus model exported by NanoDet.
+ */
 class FASTDEPLOY_DECL NanoDetPlus : public FastDeployModel {
  public:
+  /** \brief  Set path of model file and the configuration of runtime.
+   *
+   * \param[in] model_file Path of model file, e.g ./nanodet_plus_320.onnx
+   * \param[in] params_file Path of parameter file, e.g ppyoloe/model.pdiparams, if the model format is ONNX, this parameter will be ignored
+   * \param[in] custom_option RuntimeOption for inference, the default will use cpu, and choose the backend defined in "valid_cpu_backends"
+   * \param[in] model_format Model format of the loaded model, default is ONNX format
+   */
   NanoDetPlus(const std::string& model_file,
               const std::string& params_file = "",
               const RuntimeOption& custom_option = RuntimeOption(),
               const ModelFormat& model_format = ModelFormat::ONNX);
-
+  /// Get model's name
   std::string ModelName() const { return "nanodet"; }
 
-
+  /** \brief Predict the detection result for an input image
+   *
+   * \param[in] im The input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
+   * \param[in] result The output detection result will be writen to this structure
+   * \param[in] conf_threshold confidence threashold for postprocessing, default is 0.35
+   * \param[in] nms_iou_threshold iou threashold for NMS, default is 0.5
+   * \return true if the prediction successed, otherwise false
+   */
   virtual bool Predict(cv::Mat* im, DetectionResult* result,
                        float conf_threshold = 0.35f,
                        float nms_iou_threshold = 0.5f);
 
-  // tuple of input size (width, height), e.g (320, 320)
+  /// tuple of input size (width, height), e.g (320, 320)
   std::vector<int> size;
-  // padding value, size should be same with Channels
+  /// padding value, size should be the same as channels
   std::vector<float> padding_value;
-  // keep aspect ratio or not when perform resize operation.
-  // This option is set as `false` by default in NanoDet-Plus.
+  /*! @brief
+  keep aspect ratio or not when perform resize operation. This option is set as `false` by default in NanoDet-Plus
+  */
   bool keep_ratio;
-  // downsample strides for NanoDet-Plus to generate anchors, will
-  // take (8, 16, 32, 64) as default values.
+  /*! @brief
+  downsample strides for NanoDet-Plus to generate anchors, will take (8, 16, 32, 64) as default values
+  */
   std::vector<int> downsample_strides;
-  // for offseting the boxes by classes when using NMS, default 4096.
+  /// for offseting the boxes by classes when using NMS, default 4096
   float max_wh;
-  // reg_max for GFL regression, default 7
+  /// reg_max for GFL regression, default 7
   int reg_max;
 
  private:

diff --git a/fastdeploy/vision/detection/contrib/scaledyolov4.h b/fastdeploy/vision/detection/contrib/scaledyolov4.h
@@ -20,35 +20,53 @@
 namespace fastdeploy {
 namespace vision {
 namespace detection {
-
+/*! @brief ScaledYOLOv4 model object used when to load a ScaledYOLOv4 model exported by ScaledYOLOv4.
+ */
 class FASTDEPLOY_DECL ScaledYOLOv4 : public FastDeployModel {
  public:
+  /** \brief  Set path of model file and the configuration of runtime.
+   *
+   * \param[in] model_file Path of model file, e.g ./scaled_yolov4.onnx
+   * \param[in] params_file Path of parameter file, e.g ppyoloe/model.pdiparams, if the model format is ONNX, this parameter will be ignored
+   * \param[in] custom_option RuntimeOption for inference, the default will use cpu, and choose the backend defined in "valid_cpu_backends"
+   * \param[in] model_format Model format of the loaded model, default is ONNX format
+   */
+
   ScaledYOLOv4(const std::string& model_file,
                const std::string& params_file = "",
                const RuntimeOption& custom_option = RuntimeOption(),
                const ModelFormat& model_format = ModelFormat::ONNX);
 
   virtual std::string ModelName() const { return "ScaledYOLOv4"; }
-
+  /** \brief Predict the detection result for an input image
+   *
+   * \param[in] im The input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
+   * \param[in] result The output detection result will be writen to this structure
+   * \param[in] conf_threshold confidence threashold for postprocessing, default is 0.25
+   * \param[in] nms_iou_threshold iou threashold for NMS, default is 0.5
+   * \return true if the prediction successed, otherwise false
+   */
   virtual bool Predict(cv::Mat* im, DetectionResult* result,
                        float conf_threshold = 0.25,
                        float nms_iou_threshold = 0.5);
 
-  // tuple of (width, height)
+  /// tuple of (width, height)
   std::vector<int> size;
-  // padding value, size should be same with Channels
+  /// padding value, size should be the same as channels
   std::vector<float> padding_value;
-  // only pad to the minimum rectange which height and width is times of stride
+  /// only pad to the minimum rectange which height and width is times of stride
   bool is_mini_pad;
-  // while is_mini_pad = false and is_no_pad = true, will resize the image to
-  // the set size
+  /*! @brief
+  while is_mini_pad = false and is_no_pad = true, will resize the image to the set size
+  */
   bool is_no_pad;
-  // if is_scale_up is false, the input image only can be zoom out, the maximum
-  // resize scale cannot exceed 1.0
+  /*! @brief
+  if is_scale_up is false, the input image only can be zoom out, the maximum resize scale cannot exceed 1.0
+  */
   bool is_scale_up;
-  // padding stride, for is_mini_pad
+  /// padding stride, for is_mini_pad
   int stride;
-  // for offseting the boxes by classes when using NMS
+  /// for offseting the boxes by classes when using NMS
   float max_wh;
 
  private:
@@ -70,7 +88,7 @@ class FASTDEPLOY_DECL ScaledYOLOv4 : public FastDeployModel {
   // or not.)
   // while is_dynamic_shape if 'false', is_mini_pad will force 'false'. This
   // value will
-  // auto check by fastdeploy after the internal Runtime already initialized.
+  // auto check by fastdeploy after the internal Runtime already initialized
   bool is_dynamic_input_;
 };
 }  // namespace detection

diff --git a/fastdeploy/vision/detection/contrib/yolor.h b/fastdeploy/vision/detection/contrib/yolor.h
@@ -1,4 +1,4 @@
-// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.  //NOLINT
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -20,34 +20,51 @@
 namespace fastdeploy {
 namespace vision {
 namespace detection {
-
+/*! @brief YOLOR model object used when to load a YOLOR model exported by YOLOR.
+ */
 class FASTDEPLOY_DECL YOLOR : public FastDeployModel {
  public:
+  /** \brief  Set path of model file and the configuration of runtime.
+   *
+   * \param[in] model_file Path of model file, e.g ./yolor.onnx
+   * \param[in] params_file Path of parameter file, e.g ppyoloe/model.pdiparams, if the model format is ONNX, this parameter will be ignored
+   * \param[in] custom_option RuntimeOption for inference, the default will use cpu, and choose the backend defined in "valid_cpu_backends"
+   * \param[in] model_format Model format of the loaded model, default is ONNX format
+   */
   YOLOR(const std::string& model_file, const std::string& params_file = "",
         const RuntimeOption& custom_option = RuntimeOption(),
         const ModelFormat& model_format = ModelFormat::ONNX);
 
   virtual std::string ModelName() const { return "YOLOR"; }
-
+  /** \brief Predict the detection result for an input image
+   *
+   * \param[in] im The input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
+   * \param[in] result The output detection result will be writen to this structure
+   * \param[in] conf_threshold confidence threashold for postprocessing, default is 0.25
+   * \param[in] nms_iou_threshold iou threashold for NMS, default is 0.5
+   * \return true if the prediction successed, otherwise false
+   */
   virtual bool Predict(cv::Mat* im, DetectionResult* result,
                        float conf_threshold = 0.25,
                        float nms_iou_threshold = 0.5);
 
-  // tuple of (width, height)
+  /// tuple of (width, height)
   std::vector<int> size;
-  // padding value, size should be same with Channels
+  /// padding value, size should be the same as channels
   std::vector<float> padding_value;
-  // only pad to the minimum rectange which height and width is times of stride
+  /// only pad to the minimum rectange which height and width is times of stride
   bool is_mini_pad;
-  // while is_mini_pad = false and is_no_pad = true, will resize the image to
-  // the set size
+  /*! @brief
+  while is_mini_pad = false and is_no_pad = true, will resize the image to the set size
+  */
   bool is_no_pad;
-  // if is_scale_up is false, the input image only can be zoom out, the maximum
-  // resize scale cannot exceed 1.0
+  /*! @brief
+  if is_scale_up is false, the input image only can be zoom out, the maximum resize scale cannot exceed 1.0
+  */
   bool is_scale_up;
-  // padding stride, for is_mini_pad
+  /// padding stride, for is_mini_pad
   int stride;
-  // for offseting the boxes by classes when using NMS
+  /// for offseting the boxes by classes when using NMS
   float max_wh;
 
  private:
@@ -72,6 +89,7 @@ class FASTDEPLOY_DECL YOLOR : public FastDeployModel {
   // auto check by fastdeploy after the internal Runtime already initialized.
   bool is_dynamic_input_;
 };
+
 }  // namespace detection
 }  // namespace vision
 }  // namespace fastdeploy
diff --git a/fastdeploy/vision/detection/contrib/yolov5.h b/fastdeploy/vision/detection/contrib/yolov5.h
@@ -1,4 +1,4 @@
-// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.  //NOLINT
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -21,17 +21,32 @@
 namespace fastdeploy {
 namespace vision {
 namespace detection {
-
+/*! @brief YOLOv5 model object used when to load a YOLOv5 model exported by YOLOv5.
+ */
 class FASTDEPLOY_DECL YOLOv5 : public FastDeployModel {
  public:
+  /** \brief  Set path of model file and the configuration of runtime.
+   *
+   * \param[in] model_file Path of model file, e.g ./yolov5.onnx
+   * \param[in] params_file Path of parameter file, e.g ppyoloe/model.pdiparams, if the model format is ONNX, this parameter will be ignored
+   * \param[in] custom_option RuntimeOption for inference, the default will use cpu, and choose the backend defined in "valid_cpu_backends"
+   * \param[in] model_format Model format of the loaded model, default is ONNX format
+   */
   YOLOv5(const std::string& model_file, const std::string& params_file = "",
          const RuntimeOption& custom_option = RuntimeOption(),
          const ModelFormat& model_format = ModelFormat::ONNX);
 
   ~YOLOv5();
 
   std::string ModelName() const { return "yolov5"; }
-
+  /** \brief Predict the detection result for an input image
+   *
+   * \param[in] im The input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
+   * \param[in] result The output detection result will be writen to this structure
+   * \param[in] conf_threshold confidence threashold for postprocessing, default is 0.25
+   * \param[in] nms_iou_threshold iou threashold for NMS, default is 0.5
+   * \return true if the prediction successed, otherwise false
+   */
   virtual bool Predict(cv::Mat* im, DetectionResult* result,
                        float conf_threshold = 0.25,
                        float nms_iou_threshold = 0.5);
@@ -62,23 +77,25 @@ class FASTDEPLOY_DECL YOLOv5 : public FastDeployModel {
       float conf_threshold, float nms_iou_threshold, bool multi_label,
       float max_wh = 7680.0);
 
-  // tuple of (width, height)
+  /// tuple of (width, height)
   std::vector<int> size_;
-  // padding value, size should be same with Channels
+  /// padding value, size should be the same as channels
   std::vector<float> padding_value_;
-  // only pad to the minimum rectange which height and width is times of stride
+  /// only pad to the minimum rectange which height and width is times of stride
   bool is_mini_pad_;
-  // while is_mini_pad = false and is_no_pad = true, will resize the image to
-  // the set size
+  /*! @brief
+  while is_mini_pad = false and is_no_pad = true, will resize the image to the set size
+  */
   bool is_no_pad_;
-  // if is_scale_up is false, the input image only can be zoom out, the maximum
-  // resize scale cannot exceed 1.0
+  /*! @brief
+  if is_scale_up is false, the input image only can be zoom out, the maximum resize scale cannot exceed 1.0
+  */
   bool is_scale_up_;
-  // padding stride, for is_mini_pad
+  /// padding stride, for is_mini_pad
   int stride_;
-  // for offseting the boxes by classes when using NMS
+  /// for offseting the boxes by classes when using NMS
   float max_wh_;
-  // for different strategies to get boxes when postprocessing
+  /// for different strategies to get boxes when postprocessing
   bool multi_label_;
 
  private: