Files
FastDeploy/examples/vision/facedet/blazeface/cpp/README_CN.md
CoolCola 42d14e7119 [Model] Support BlazeFace Model (#1172)
* fit yolov7face file path

* TODO:添加yolov7facePython接口Predict

* resolve yolov7face.py

* resolve yolov7face.py

* resolve yolov7face.py

* add yolov7face example readme file

* [Doc] fix yolov7face example readme file

* [Doc]fix yolov7face example readme file

* support BlazeFace

* add blazeface readme file

* fix review problem

* fix code style error

* fix review problem

* fix review problem

* fix head file problem

* fix review problem

* fix review problem

* fix readme file problem

* add English readme file

* fix English readme file
2023-02-06 14:24:12 +08:00

3.1 KiB
Raw Blame History

English | 简体中文

BlazeFace C++部署示例

本目录下提供infer.cc快速完成BlazeFace在CPU/GPU部署的示例。

在部署前,需确认以下两个步骤

以Linux上CPU推理为例在本目录执行如下命令即可完成编译测试

mkdir build
cd build
# 下载FastDeploy预编译库用户可在上文提到的`FastDeploy预编译库`中自行选择合适的版本使用
wget https://bj.bcebos.com/fastdeploy/release/cpp/fastdeploy-linux-x64-x.x.x.tgz # x.x.x >= 1.0.4
tar xvf fastdeploy-linux-x64-x.x.x.tgz # x.x.x >= 1.0.4
cmake .. -DFASTDEPLOY_INSTALL_DIR=${PWD}/fastdeploy-linux-x64-x.x.x # x.x.x >= 1.0.4
make -j

#下载官方转换好的BlazeFace模型文件和测试图片
wget https://raw.githubusercontent.com/DefTruth/lite.ai.toolkit/main/examples/lite/resources/test_lite_face_detector_3.jpg
wget https://bj.bcebos.com/paddlehub/fastdeploy/blzeface-1000e.tgz

#使用blazeface-1000e模型
# CPU推理
./infer_demo blazeface-1000e/ test_lite_face_detector_3.jpg 0
# GPU推理
./infer_demo blazeface-1000e/ test_lite_face_detector_3.jpg 1

运行完成可视化结果如下图所示

<img width="640" src="https://user-images.githubusercontent.com/49013063/206170111-843febb6-67d6-4c46-a121-d87d003bba21.jpg">

以上命令只适用于Linux或MacOS, Windows下SDK的使用方式请参考:  
- [如何在Windows中使用FastDeploy C++ SDK](../../../../../docs/cn/faq/use_sdk_on_windows.md)

## BlazeFace C++接口

### BlazeFace类

```c++
fastdeploy::vision::facedet::BlazeFace(
        const string& model_file,
        const string& params_file = "",
        const string& config_file = "",
        const RuntimeOption& runtime_option = RuntimeOption(),
        const ModelFormat& model_format = ModelFormat::PADDLE)

BlazeFace模型加载和初始化其中model_file为导出的PADDLE模型格式。

参数

  • model_file(str): 模型文件路径
  • params_file(str): 参数文件路径当模型格式为ONNX时此参数传入空字符串即可
  • config_file(str): 配置文件路径当模型格式为ONNX时此参数传入空字符串即可
  • runtime_option(RuntimeOption): 后端推理配置默认为None即采用默认配置
  • model_format(ModelFormat): 模型格式默认为PADDLE格式

Predict函数

BlazeFace::Predict(cv::Mat& im, FaceDetectionResult* result)

模型预测接口,输入图像直接输出检测结果。

参数

  • im: 输入图像注意需为HWCBGR格式
  • result: 检测结果,包括检测框,各个框的置信度, FaceDetectionResult说明参考视觉模型预测结果