[Model] Support New PaddleDetection Models (#959)

* Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload * Add files via upload Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com>
2025-10-05 16:48:03 +08:00 · 2023-01-04 10:34:21 +08:00
parent 96ca92ce34
commit 9b108c3fa5
23 changed files with 1615 additions and 3 deletions
--- a/examples/vision/detection/paddledetection/README.md
+++ b/examples/vision/detection/paddledetection/README.md
@@ -20,6 +20,15 @@
 - [YOLOv6系列模型](https://github.com/PaddlePaddle/PaddleYOLO/tree/release/2.5/configs/yolov6)
 - [YOLOv7系列模型](https://github.com/PaddlePaddle/PaddleYOLO/tree/release/2.5/configs/yolov7)
 - [RTMDet系列模型](https://github.com/PaddlePaddle/PaddleYOLO/tree/release/2.5/configs/rtmdet)
+- [CascadeRCNN系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/cascade_rcnn)
+- [PSSDet系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/rcnn_enhance)
+- [RetinaNet系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/retinanet)
+- [PPYOLOESOD系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/smalldet)
+- [FCOS系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/fcos)
+- [TTFNet系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/ttfnet)
+- [TOOD系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/tood)
+- [GFL系列模型](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/gfl)
+

 ## 导出部署模型

@@ -59,6 +68,16 @@
 | [yolov6_s_400e_coco](https://bj.bcebos.com/paddlehub/fastdeploy/yolov6_s_400e_coco.tgz) | 68M | Box AP 43.4%|  |
 | [yolov7_l_300e_coco](https://bj.bcebos.com/paddlehub/fastdeploy/yolov7_l_300e_coco.tgz) | 145M | Box AP 51.0%|  |
 | [yolov7_x_300e_coco](https://bj.bcebos.com/paddlehub/fastdeploy/yolov7_x_300e_coco.tgz) | 277M | Box AP 53.0%|  |
+| [cascade_rcnn_r50_fpn_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/cascade_rcnn_r50_fpn_1x_coco.tgz) | 271M | Box AP 41.1%|  暂不支持TensorRT、ORT |
+| [cascade_rcnn_r50_vd_fpn_ssld_2x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/cascade_rcnn_r50_vd_fpn_ssld_2x_coco.tgz) | 271M | Box AP 45.0%|  暂不支持TensorRT、ORT |
+| [faster_rcnn_enhance_3x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/faster_rcnn_enhance_3x_coco.tgz) | 119M | Box AP 41.5%|  暂不支持TensorRT、ORT |
+| [fcos_r50_fpn_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/fcos_r50_fpn_1x_coco.tgz) | 129M | Box AP 39.6%|  暂不支持TensorRT |
+| [gfl_r50_fpn_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/gfl_r50_fpn_1x_coco.tgz) | 128M | Box AP 41.0%|  暂不支持TensorRT |
+| [ppyoloe_crn_l_80e_sliced_visdrone_640_025](https://bj.bcebos.com/paddlehub/fastdeploy/ppyoloe_crn_l_80e_sliced_visdrone_640_025.tgz) | 200M | Box AP 31.9%|  |
+| [retinanet_r101_fpn_2x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/retinanet_r101_fpn_2x_coco.tgz) | 210M | Box AP 40.6%|  暂不支持TensorRT、ORT |
+| [retinanet_r50_fpn_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/retinanet_r50_fpn_1x_coco.tgz) | 136M | Box AP 37.5%|  暂不支持TensorRT、ORT |
+| [tood_r50_fpn_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/tood_r50_fpn_1x_coco.tgz) | 130M | Box AP 42.5%|  暂不支持TensorRT、ORT |
+| [ttfnet_darknet53_1x_coco](https://bj.bcebos.com/paddlehub/fastdeploy/ttfnet_darknet53_1x_coco.tgz) | 178M | Box AP 33.5%|  暂不支持TensorRT、ORT |

 ## 详细部署文档

--- a/examples/vision/detection/paddledetection/cpp/CMakeLists.txt
+++ b/examples/vision/detection/paddledetection/cpp/CMakeLists.txt
@@ -44,3 +44,27 @@ target_link_libraries(infer_yolov7_demo ${FASTDEPLOY_LIBS})

 add_executable(infer_rtmdet_demo ${PROJECT_SOURCE_DIR}/infer_rtmdet.cc)
 target_link_libraries(infer_rtmdet_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_cascadercnn_demo ${PROJECT_SOURCE_DIR}/infer_cascadercnn.cc)
+target_link_libraries(infer_cascadercnn_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_pssdet_demo ${PROJECT_SOURCE_DIR}/infer_pssdet.cc)
+target_link_libraries(infer_pssdet_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_retinanet_demo ${PROJECT_SOURCE_DIR}/infer_retinanet.cc)
+target_link_libraries(infer_retinanet_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_ppyoloesod_demo ${PROJECT_SOURCE_DIR}/infer_ppyoloesod.cc)
+target_link_libraries(infer_ppyoloesod_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_fcos_demo ${PROJECT_SOURCE_DIR}/infer_fcos.cc)
+target_link_libraries(infer_fcos_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_ttfnet_demo ${PROJECT_SOURCE_DIR}/infer_ttfnet.cc)
+target_link_libraries(infer_ttfnet_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_tood_demo ${PROJECT_SOURCE_DIR}/infer_tood.cc)
+target_link_libraries(infer_tood_demo ${FASTDEPLOY_LIBS})
+
+add_executable(infer_gfl_demo ${PROJECT_SOURCE_DIR}/infer_gfl.cc)
+target_link_libraries(infer_gfl_demo ${FASTDEPLOY_LIBS})
--- a/examples/vision/detection/paddledetection/cpp/README.md
+++ b/examples/vision/detection/paddledetection/cpp/README.md
@@ -1,6 +1,6 @@
 # PaddleDetection C++部署示例

-本目录下提供`infer_xxx.cc`快速完成PaddleDetection模型包括PPYOLOE/PicoDet/YOLOX/YOLOv3/PPYOLO/FasterRCNN/YOLOv5/YOLOv6/YOLOv7/RTMDet在CPU/GPU，以及GPU上通过TensorRT加速部署的示例。
+本目录下提供`infer_xxx.cc`快速完成PaddleDetection模型包括PPYOLOE/PicoDet/YOLOX/YOLOv3/PPYOLO/FasterRCNN/YOLOv5/YOLOv6/YOLOv7/RTMDet/CascadeRCNN/PSSDet/RetinaNet/PPYOLOESOD/FCOS/TTFNet/TOOD/GFL在CPU/GPU，以及GPU上通过TensorRT加速部署的示例。

 在部署前，需确认以下两个步骤

@@ -45,7 +45,7 @@ tar xvf ppyoloe_crn_l_300e_coco.tgz

 ### 模型类

-PaddleDetection目前支持6种模型系列，类名分别为`PPYOLOE`, `PicoDet`, `PaddleYOLOX`, `PPYOLO`, `FasterRCNN`，`SSD`,`PaddleYOLOv5`,`PaddleYOLOv6`,`PaddleYOLOv7`,`RTMDet`所有类名的构造函数和预测函数在参数上完全一致，本文档以PPYOLOE为例讲解API
+PaddleDetection目前支持6种模型系列，类名分别为`PPYOLOE`, `PicoDet`, `PaddleYOLOX`, `PPYOLO`, `FasterRCNN`，`SSD`,`PaddleYOLOv5`,`PaddleYOLOv6`,`PaddleYOLOv7`,`RTMDet`,`CascadeRCNN`,`PSSDet`,`RetinaNet`,`PPYOLOESOD`,`FCOS`,`TTFNet`,`TOOD`,`GFL`所有类名的构造函数和预测函数在参数上完全一致，本文档以PPYOLOE为例讲解API
 ```c++
 fastdeploy::vision::detection::PPYOLOE(
        const string& model_file,
--- a/examples/vision/detection/paddledetection/cpp/infer_cascadercnn.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_cascadercnn.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::CascadeRCNN(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::CascadeRCNN(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  } 
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_fcos.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_fcos.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::FCOS(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::FCOS(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_gfl.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_gfl.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::GFL(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::GFL(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  } 
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_ppyoloesod.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_ppyoloesod.cc
@@ -0,0 +1,127 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::PPYOLOESOD(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::PPYOLOESOD(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void TrtInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  option.UseTrtBackend();
+  auto model = fastdeploy::vision::detection::PPYOLOESOD(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 2) {
+    TrtInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_pssdet.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_pssdet.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::PSSDet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::PSSDet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_retinanet.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_retinanet.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::RetinaNet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::RetinaNet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_tood.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_tood.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::TOOD(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::TOOD(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/cpp/infer_ttfnet.cc
+++ b/examples/vision/detection/paddledetection/cpp/infer_ttfnet.cc
@@ -0,0 +1,96 @@
+// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "fastdeploy/vision.h"
+
+#ifdef WIN32
+const char sep = '\\';
+#else
+const char sep = '/';
+#endif
+
+void CpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+  auto option = fastdeploy::RuntimeOption();
+  option.UseCpu();
+  auto model = fastdeploy::vision::detection::TTFNet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_dir, const std::string& image_file) {
+  auto model_file = model_dir + sep + "model.pdmodel";
+  auto params_file = model_dir + sep + "model.pdiparams";
+  auto config_file = model_dir + sep + "infer_cfg.yml";
+
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::TTFNet(model_file, params_file,
+                                                      config_file, option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  std::cout << res.Str() << std::endl;
+  auto vis_im = fastdeploy::vision::VisDetection(im, res, 0.5);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout
+        << "Usage: infer_demo path/to/model_dir path/to/image run_option, "
+           "e.g ./infer_model ./ppyoloe_model_dir ./test.jpeg 0"
+        << std::endl;
+    std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
+                 "with gpu; 2: run with gpu and use tensorrt backend."
+              << std::endl;
+    return -1;
+  }
+
+  if (std::atoi(argv[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/paddledetection/python/README.md
+++ b/examples/vision/detection/paddledetection/python/README.md
@@ -49,6 +49,14 @@ fastdeploy.vision.detection.PaddleYOLOv5(model_file, params_file, config_file, r
 fastdeploy.vision.detection.PaddleYOLOv6(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
 fastdeploy.vision.detection.PaddleYOLOv7(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
 fastdeploy.vision.detection.RTMDet(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.CascadeRCNN(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.PSSDet(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.RetinaNet(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.PPYOLOESOD(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.FCOS(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.TTFNet(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.TOOD(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+fastdeploy.vision.detection.GFL(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
 ```

 PaddleDetection模型加载和初始化，其中model_file， params_file为导出的Paddle部署模型格式, config_file为PaddleDetection同时导出的部署配置yaml文件
--- a/examples/vision/detection/paddledetection/python/infer_cascadercnn.py
+++ b/examples/vision/detection/paddledetection/python/infer_cascadercnn.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.CascadeRCNN(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_fcos.py
+++ b/examples/vision/detection/paddledetection/python/infer_fcos.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.FCOS(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_gfl.py
+++ b/examples/vision/detection/paddledetection/python/infer_gfl.py
@@ -0,0 +1,51 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.") 
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.GFL(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im.copy())
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_ppyoloesod.py
+++ b/examples/vision/detection/paddledetection/python/infer_ppyoloesod.py
@@ -0,0 +1,59 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    parser.add_argument(
+        "--use_trt",
+        type=ast.literal_eval,
+        default=False,
+        help="Wether to use tensorrt.")    
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+
+    if args.use_trt:
+        option.use_trt_backend()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.PPYOLOESOD(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im.copy())
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_pssdet.py
+++ b/examples/vision/detection/paddledetection/python/infer_pssdet.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.PSSDet(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_retinanet.py
+++ b/examples/vision/detection/paddledetection/python/infer_retinanet.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.RetinaNet(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_tood.py
+++ b/examples/vision/detection/paddledetection/python/infer_tood.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.TOOD(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/examples/vision/detection/paddledetection/python/infer_ttfnet.py
+++ b/examples/vision/detection/paddledetection/python/infer_ttfnet.py
@@ -0,0 +1,50 @@
+import fastdeploy as fd
+import cv2
+import os
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_dir",
+        required=True,
+        help="Path of PaddleDetection model directory")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model_dir, "model.pdmodel")
+params_file = os.path.join(args.model_dir, "model.pdiparams")
+config_file = os.path.join(args.model_dir, "infer_cfg.yml")
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.TTFNet(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result, score_threshold=0.5)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")
--- a/fastdeploy/vision/detection/ppdet/model.h
+++ b/fastdeploy/vision/detection/ppdet/model.h
@@ -260,6 +260,134 @@ class FASTDEPLOY_DECL RTMDet : public PPDetBase {
  virtual std::string ModelName() const { return "PaddleDetection/RTMDet"; }
 };

+class FASTDEPLOY_DECL CascadeRCNN : public PPDetBase {
+ public:
+  CascadeRCNN(const std::string& model_file, const std::string& params_file,
+              const std::string& config_file,
+              const RuntimeOption& custom_option = RuntimeOption(),
+              const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/CascadeRCNN"; }
+};
+
+class FASTDEPLOY_DECL PSSDet : public PPDetBase {
+ public:
+  PSSDet(const std::string& model_file, const std::string& params_file,
+         const std::string& config_file,
+         const RuntimeOption& custom_option = RuntimeOption(),
+         const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/PSSDet"; }
+};
+
+class FASTDEPLOY_DECL RetinaNet : public PPDetBase {
+ public:
+  RetinaNet(const std::string& model_file, const std::string& params_file,
+            const std::string& config_file,
+            const RuntimeOption& custom_option = RuntimeOption(),
+            const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/RetinaNet"; }
+};
+
+class FASTDEPLOY_DECL PPYOLOESOD : public PPDetBase {
+ public:
+  PPYOLOESOD(const std::string& model_file, const std::string& params_file,
+             const std::string& config_file,
+             const RuntimeOption& custom_option = RuntimeOption(),
+             const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::ORT, Backend::PDINFER};
+    valid_gpu_backends = {Backend::ORT, Backend::PDINFER, Backend::TRT};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/PPYOLOESOD"; }
+};
+
+class FASTDEPLOY_DECL FCOS : public PPDetBase {
+ public:
+  FCOS(const std::string& model_file, const std::string& params_file,
+       const std::string& config_file,
+       const RuntimeOption& custom_option = RuntimeOption(),
+       const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::ORT, Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/FCOS"; }
+};
+
+class FASTDEPLOY_DECL TTFNet : public PPDetBase {
+ public:
+  TTFNet(const std::string& model_file, const std::string& params_file,
+         const std::string& config_file,
+         const RuntimeOption& custom_option = RuntimeOption(),
+         const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/TTFNet"; }
+};
+
+class FASTDEPLOY_DECL TOOD : public PPDetBase {
+ public:
+  TOOD(const std::string& model_file, const std::string& params_file,
+       const std::string& config_file,
+       const RuntimeOption& custom_option = RuntimeOption(),
+       const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::PDINFER};
+    valid_gpu_backends = {Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/TOOD"; }
+};
+
+class FASTDEPLOY_DECL GFL : public PPDetBase {
+ public:
+  GFL(const std::string& model_file, const std::string& params_file,
+      const std::string& config_file,
+      const RuntimeOption& custom_option = RuntimeOption(),
+      const ModelFormat& model_format = ModelFormat::PADDLE)
+      : PPDetBase(model_file, params_file, config_file, custom_option,
+                model_format) {
+    valid_cpu_backends = {Backend::ORT, Backend::PDINFER};
+    valid_gpu_backends = {Backend::ORT, Backend::PDINFER};
+    initialized = Initialize();
+  }
+
+  virtual std::string ModelName() const { return "PaddleDetection/GFL"; }
+};
+
 }  // namespace detection
 }  // namespace vision
 }  // namespace fastdeploy
--- a/fastdeploy/vision/detection/ppdet/ppdet_pybind.cc
+++ b/fastdeploy/vision/detection/ppdet/ppdet_pybind.cc
@@ -137,5 +137,29 @@ void BindPPDet(pybind11::module& m) {
  pybind11::class_<vision::detection::RTMDet, vision::detection::PPDetBase>(m, "RTMDet")
      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
                          ModelFormat>());    
+  pybind11::class_<vision::detection::CascadeRCNN, vision::detection::PPDetBase>(m, "CascadeRCNN")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());    
+  pybind11::class_<vision::detection::PSSDet, vision::detection::PPDetBase>(m, "PSSDet")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::RetinaNet, vision::detection::PPDetBase>(m, "RetinaNet")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::PPYOLOESOD, vision::detection::PPDetBase>(m, "PPYOLOESOD")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::FCOS, vision::detection::PPDetBase>(m, "FCOS")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::TTFNet, vision::detection::PPDetBase>(m, "TTFNet")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::TOOD, vision::detection::PPDetBase>(m, "TOOD")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());
+  pybind11::class_<vision::detection::GFL, vision::detection::PPDetBase>(m, "GFL")
+      .def(pybind11::init<std::string, std::string, std::string, RuntimeOption,
+                          ModelFormat>());                                                                                                                                                                                              
 }
 }  // namespace fastdeploy
--- a/python/fastdeploy/vision/detection/ppdet/init.py
+++ b/python/fastdeploy/vision/detection/ppdet/init.py
@@ -513,3 +513,203 @@ class RTMDet(PPYOLOE):
            model_file, params_file, config_file, self._runtime_option,
            model_format)
        assert self.initialized, "RTMDet model initialize failed."
+
+
+class CascadeRCNN(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a CascadeRCNN model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g cascadercnn/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g cascadercnn/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "CascadeRCNN model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.CascadeRCNN(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "CascadeRCNN model initialize failed."
+
+
+class PSSDet(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a PSSDet model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g pssdet/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g pssdet/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "PSSDet model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.PSSDet(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "PSSDet model initialize failed."
+
+
+class RetinaNet(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a RetinaNet model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g retinanet/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g retinanet/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "RetinaNet model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.RetinaNet(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "RetinaNet model initialize failed."
+
+
+class PPYOLOESOD(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a PPYOLOESOD model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g ppyoloesod/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g ppyoloesod/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "PPYOLOESOD model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.PPYOLOESOD(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "PPYOLOESOD model initialize failed."
+
+
+class FCOS(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a FCOS model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g fcos/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g fcos/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "FCOS model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.FCOS(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "FCOS model initialize failed."
+
+
+class TTFNet(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a TTFNet model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g ttfnet/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g ttfnet/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "TTFNet model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.TTFNet(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "TTFNet model initialize failed."
+
+
+class TOOD(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a TOOD model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g tood/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g tood/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "TOOD model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.TOOD(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "TOOD model initialize failed."
+
+
+class GFL(PPYOLOE):
+    def __init__(self,
+                 model_file,
+                 params_file,
+                 config_file,
+                 runtime_option=None,
+                 model_format=ModelFormat.PADDLE):
+        """Load a GFL model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g gfl/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g gfl/model.pdiparams, if the model_fomat is ModelFormat.ONNX, this param will be ignored, can be set as empty string
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
+
+        super(PPYOLOE, self).__init__(runtime_option)
+
+        assert model_format == ModelFormat.PADDLE, "GFL model only support model format of ModelFormat.Paddle now."
+        self._model = C.vision.detection.GFL(
+            model_file, params_file, config_file, self._runtime_option,
+            model_format)
+        assert self.initialized, "GFL model initialize failed."