【Ubuntu18.04】Seetaface6人脸识别部署

Seetaface6人脸识别部署

一、下载源码

使用git命令

git clone --recursive https://github.com/SeetaFace6Open/index.git

二、编译工具

• make

  sudo apt install make
  

• gcc(系统自带)

  sudo apt install gcc
  

• g++(系统自带)

  sudo apt install g++
  

• cmake

  sudo apt install cmake 
  

三、编译

3.1 编译OpenRoleZoo

进入编译脚本所在文件夹

cd index/OpenRoleZoo
cd craft

执行编译脚本编译

chmod +777 build.linux.x64.sh 
./build.linux.x64.sh

出现如下报错
error: funciton in namespace 'std' does not name a template type

解决方案：
进入提示信息所给的pot.h文件中，导入头文件functional后保存文件重新编译

#include <functional>

保存后重新执行脚本文件

3.2 编译SeetaAuthorize

进入编译脚本所在文件夹

cd index/SeetaAuthorize
cd craft

执行编译脚本编译

chmod +777 build.linux.x64.sh 
./build.linux.x64.sh

3.3 编译TenniS

进入编译脚本所在文件夹

cd index/TenniS
cd craft

执行编译脚本编译

chmod +777 build.linux.x64.sh 
./build.linux.x64.sh

3.2 编译剩余模块

方法与前三个模块的编译相同

chmod +777 build.linux.x64.sh 
./build.linux.x64.sh

可以写一个shell脚本程序进行编译
执行即可编译剩余模块

cd FaceTracker6/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd SeetaEyeStateDetector/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd Landmarker/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd SeetaGenderPredictor/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd OpenRoleZoo/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd SeetaMaskDetector/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd FaceAntiSpoofingX6/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd PoseEstimator6/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd FaceBoxes/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd QualityAssessor3/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd FaceRecognizer6/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

cd SeetaAgePredictor/craft
chmod +777 build.linux.x64.sh 
./build.linux.x64.sh
cd ../..

四、运行demo

4.1 修改lib路径

如下图

4.2 build project

build成功后会生成文件夹

建立文件夹models存放模型

从链接（https://pan.baidu.com/s/1LlXe2-YsUxQMe-MLzhQ2Aw 提取码：ngne）下载模型并解压到models文件夹下。

4.2 运行 Run！

菜单界面：

人脸数据界面：

相机界面：

设置界面：