人脸检测dlib, mtcnnx对比，FDDB测试对比

素颜马尾好姑娘i 2022-05-20 07:06 413阅读 0赞

问题来源 [How does MTCNN perform vs DLIB for face detection?][How does MTCNN perform vs DLIB for face detection]

前面一直做人脸检测相关内容，然后对比了下dib以及MTCNN的人脸检测效果主要是速度，以及FDDB准确率。最后给出生成FDDB测试文件的C++代码。

--------------------

#### FDDB 测试结果 ####

注本文的MTCNN效果检测准确率不是最优的，最优的在FDDB上可达95%，测试效果如下：

![这里写图片描述][70]

可以看到三种方法：

1.  MTCNN 大概90%
2.  dlib 大概 77%
3.  opencv 大概 62%

dlib的作者非要说我的测试有问题，如果谁感兴趣可以使用dlib测试下FDDB的结果。

--------------------

#### 速度 ####

在CPU和GPU模式下，对于三种不同尺寸的图片，运行一千次测试平均的时效：

###### CPU模式 ######

MTCNN（既检测人脸又做landmark）：

![这里写图片描述][70 1]

dlib (仅仅检测人脸):

![这里写图片描述][70 2]

###### GPU模式 ######

MTCNN（既检测人脸又做landmark）：

![这里写图片描述][70 3]

dlib (仅仅检测人脸):

![这里写图片描述][70 4]

可以看到：

1.  在检测精度上MTCNN显然好于dlib
2.  无论是CPU还是GPU模型下MTCNN的检测数度都好于dlib，而且dlib还做了人脸的landmark

--------------------

dlib c++生成FDDB结果代码如下（至于怎么使用FDDB测试可见[前面blog，有py实现][blog_py]）或者[我的stackoverflow回答][stackoverflow]：

#include <iostream>
    #include <dlib/dnn.h>
    #include <dlib/data_io.h>
    #include <dlib/image_processing.h>
    #include <dlib/gui_widgets.h>
    using namespace std;
    using namespace dlib;
    
    // ----------------------------------------------------------------------------------------
    
    template <long num_filters, typename SUBNET> using con5d = con<num_filters,5,5,2,2,SUBNET>;
    template <long num_filters, typename SUBNET> using con5  = con<num_filters,5,5,1,1,SUBNET>;
    
    template <typename SUBNET> using downsampler  = relu<affine<con5d<32, relu<affine<con5d<32, relu<affine<con5d<16,SUBNET>>>>>>>>>;
    template <typename SUBNET> using rcon5  = relu<affine<con5<45,SUBNET>>>;
    
    using net_type = loss_mmod<con<1,9,9,1,1,rcon5<rcon5<rcon5<downsampler<input_rgb_image_pyramid<pyramid_down<6>>>>>>>>;
    
    // ----------------------------------------------------------------------------------------
    
    void getAllImgPaths(const std::string& file, std::vector<std::string>& vecPaths){
    
        std::fstream fStream(file);
        std::string sLine;
    
        while (std::getline(fStream, sLine)){
            if (sLine.size() > 0){
                vecPaths.emplace_back(sLine);
            }
        }
    
        fStream.close();
    }
    
    void writeStrVecToFile(const std::string& file, const std::vector<std::string>& vecStr){
        std::ofstream fout(file);
        for (auto const& x:vecStr){
            fout<<x<<'\n';
        }
    
        fout.close();
    }
    
    
    
    int main(){
    
        std::string fPath = "/home/xy/face_sample/evaluation/compareROC/FDDB-folds/filePath.txt";
        std::vector<std::string> vecImgPaths;
    
        getAllImgPaths(fPath, vecImgPaths);
    
        std::string imgBaseDir = "/home/xy/face_sample/evaluation/compareROC/originalPics/";
        std::vector<std::string> vecDetRet;
    
        string model_path = "/home/xy/anaconda2/lib/python2.7/site-packages/face_recognition_models/models/mmod_human_face_detector.dat";
        net_type net;
        deserialize(model_path) >> net;
    
        for (auto const& img_name:vecImgPaths){
            std::string imgFullPath = imgBaseDir + img_name + ".jpg";
    
            matrix<rgb_pixel> img;
            load_image(img, imgFullPath);
    
            auto dets = net(img);
            vecDetRet.push_back(img_name);
            vecDetRet.push_back(std::to_string(dets.size()));
    
            for (auto det:dets){
    
                using std::to_string;
    
                // sFaceInfo like 49 55 193 193 0.999784
                std::string sFaceInfo = to_string(det.rect.left()) + " " + to_string(det.rect.top()) + " " +
                                        to_string(det.rect.width()) + " " + to_string(det.rect.height()) + " " + to_string(1);
    
                std::cout<<sFaceInfo<<std::endl;
                vecDetRet.push_back(sFaceInfo);
    
            }
    
        }
    
        // write face detect result to txt file for fddb compare
        std::string fddbTxtPath = "fddb_ret.txt";
        writeStrVecToFile(fddbTxtPath, vecDetRet);
    }

[How does MTCNN perform vs DLIB for face detection]: https://stackoverflow.com/a/51054158/6329006
[70]: /images/20220520/06e6366741f54ad9b46ceebd7a4475bd.png
[70 1]: /images/20220520/f20d6647fa864d8bb4f70064c5071613.png
[70 2]: /images/20220520/77e90c68402049c587781187759e1148.png
[70 3]: /images/20220520/bee932464aee49cf8691b3b1eace205d.png
[70 4]: /images/20220520/a625448d2b9149b3aa174b6502d0ef25.png
[blog_py]: https://blog.csdn.net/haluoluo211/article/details/80698000
[stackoverflow]: https://stackoverflow.com/a/50852575/6329006