A. 194 landmark points

1. 1st data

gopi jayaraman uploaded the shape predictor .dat file for 194 facelandmarks (helen dataset)

shape_predictor_194_face_landmarks.dat 파일 있는 주소.

You can download from here:

https://www.dropbox.com/sh/t5h024w0xkedq0j/AABS3GprqIvb_PwqeHOn2dxNa?dl=0

you can use it with dlib.

you can check the above address in following address

아래 주소에 어떤 사람이 올려놨음.

https://stackoverflow.com/questions/36711905/dlib-train-shape-predictor-ex-cpp?rq=1

2. 2nd data

spHelen194.dat 파일 있는 주소.

You can download from here:

https://wisimage.blogspot.fr/2018/05/dlib-shape-predictor-models.html

3. 3rd data

helen-dataset.dat 파일 있는 주소.

You can download from here:

https://qiita.com/kekeho/items/0b2d4ed5192a4c90a0ac 

https://drive.google.com/open?id=1qMAc4xqnwejje8sK3kqgwrgUgJYFFkbH

B. 81 landmark points 

shape_predictor_81_face_landmarks.dat

https://github.com/codeniko/shape_predictor_81_face_landmarks

facelandmarks, dlib, face feature points

How can I convert a dlib::array2d to gray Image? The gray Image should be array2d or not? I never see a good dlib documentation.

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

dlib::array2d is an image already, you can use it for any dlib's image functions

load image:

dlib::array2d<dlib::rgb_pixel> img_rgb;
dlib::load_image(img_rgb, "test_image.jpg");

convert to greyscale:

dlib::array2d<unsigned char> img_gray;
dlib::assign_image(img_gray, img_rgb);

converto to OpenCV Image (cv::Mat):

#include <dlib/opencv.h>
#include <opencv2/opencv.hpp>
cv::Mat img = dlib::toMat(img_gray);

get image from OpenCV:

#include <dlib/opencv.h>
#include <opencv2/opencv.hpp>
cv::Mat img = cv::imread("test_image.jpg")
dlib::cv_image<rgb_pixel> dlib_img(img); // only stores pointer, no deep copy

Documentation is here. There are a lot of well-documented examples, you should start from them. Special example about array2d

http://stackoverflow.com/questions/36908402/dlib-training-shape-predictor-for-194-landmarks-helen-dataset

DLIB : Training Shape_predictor for 194 landmarks (helen dataset)

I am training DLIB's shape_predictor for 194 face landmarks using helen dataset which is used to detect face landmarks through face_landmark_detection_ex.cpp of dlib library.

Now it gave me an sp.dat binary file of around 45 MB which is less compared to file given (http://sourceforge.net/projects/dclib/files/dlib/v18.10/shape_predictor_68_face_landmarks.dat.bz2) for 68 face landmarks. In training

• Mean training error : 0.0203811
• Mean testing error : 0.0204511

and when I used trained data to get face landmarks position, IN result I got..

which are very deviated from the result got from 68 landmarks

68 landmark image:

Why?

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

Ok, looks like you haven't read the code comments (?):

shape_predictor_trainer trainer;
// This algorithm has a bunch of parameters you can mess with.  The
// documentation for the shape_predictor_trainer explains all of them.
// You should also read Kazemi's paper which explains all the parameters
// in great detail.  However, here I'm just setting three of them
// differently than their default values.  I'm doing this because we
// have a very small dataset.  In particular, setting the oversampling
// to a high amount (300) effectively boosts the training set size, so
// that helps this example.
trainer.set_oversampling_amount(300);
// I'm also reducing the capacity of the model by explicitly increasing
// the regularization (making nu smaller) and by using trees with
// smaller depths.  
trainer.set_nu(0.05);
trainer.set_tree_depth(2);

Have a look at the Kazemi paper, ctrl-f the string 'parameter' and have a read...

https://forum.qt.io/topic/69007/compile-dlib-on-windows

I try to use dlib in Qt project on Windows. After downloading I did this in the dlib root:

cd examples
mkdir build
cd build
cmake .. -G"Visual Studio 14 2015 Win64" 
cmake --build . --config Release

And this(again in dlib root):

mkdir build
cd build
cmake .. -G"Visual Studio 14 2015 Win64" -DCMAKE_INSTALL_PREFIX=D:\dlib_build
cmake --build . --config Release --target install

My .pro file:

QT += core
QT -= gui

CONFIG += c++11

TARGET = dlibWin2
CONFIG += console
CONFIG -= app_bundle

TEMPLATE = app

SOURCES += main.cpp

INCLUDEPATH += "D:\dlib_build\include"
LIBS += -L"D:\dlib_build\lib" -ldlib
QMAKE_CXXFLAGS_RELEASE += /arch:AVX
QMAKE_CXXFLAGS += -DDLIB_JPEG_SUPPORT

main.cpp:

#include <QCoreApplication>
#include <dlib/image_processing/frontal_face_detector.h>
#include <dlib/image_processing/render_face_detections.h>
#include <dlib/image_processing.h>
#include <dlib/gui_widgets.h>
#include <dlib/image_io.h>
#include <iostream>


using namespace dlib;
using namespace std;
int main(int argc, char *argv[])
{
    QCoreApplication a(argc, argv);


    try
    {
        frontal_face_detector detector = get_frontal_face_detector();
    }
    catch (exception& e)
    {
        cout << "\nexception thrown!" << endl;
        cout << e.what() << endl;
    }
    return a.exec();
}

Compilation output for MSVC2015 64bit Release:

D:\dlib_build\include\dlib\config.h:19: Warning: C4005: DLIB_JPEG_SUPPORT

Runtime output for MSVC2015 64bit Release:

The program has unexpectedly finished... Exited with code -1073741795

Please note that I did this after Windows reinstalling, and before this I got absolutely the same issue.

How can I solve this or how can I use dlib in Qt on Windows?

요약

0. Visual Studio Express 2015버전을 설치한다. face detection 또는 facelandmarks detection 이용하려면 opencv도 설치해야한다. (matlab으로 lbp뽑으려면 vlfeat도 설치해야. vlfeat도 업데이트되어 20180328현재 다시 새버전으로 설치. vlfeat를 쓰려면 mex파일로 만들어야하는데, 이 때 윈도우용 mex컴파일러가 필요하다. mex컴파일러 설치방법은 https://www.mathworks.com/matlabcentral/answers/233850-how-can-i-install-sdk-7-1-on-windows-10 참조)

1. dlib 파일 (예:dlib-19.10)과 cmake 파일 (예:cmake-3.7.0-rc1-win32-x86)을 다운받아 압축을 푼다.

2. 윈도우 커맨드창을 열고 dlib-19.10\examples 경로로 이동한 후 bulid폴더를 만든다.

[D:\Research\dlib\dlib-19.10\examples>mkdir build]

3. build폴더로 이동한다. [D:\Research\dlib\dlib-19.10\examples>cd build]

4. 아래와 같이 [cmake.exe의절대경로 -G "비주얼스튜디오버전" ..]를 입력한다. 가장 뒤에 ' ..'를 꼭 써줘야한다.

[D:\Research\dlib\cmake-3.7.0-rc1-win32-x86\bin\cmake.exe -G "Visual Studio 14 2015 Win64" ..]

5. cmake로 컴파일한다. 시간이 5분정도? 걸린다.

[D:\Research\dlib\dlib-19.10\examples\build>D:\Research\dlib\cmake-3.7.0-rc1-win32-x86\bin\cmake.exe --build . --config Release]

컴파일이 끝나고 가장 끝에 '경고 3개, 오류 0개'라고 뜨면 성공한 것이다.

아래 예제캡춰에서 4번식 가장 뒤에 ' ..'를 꼭 써줘야한다.

Windows / Visual Studio 환경에 dlib의 도입 방법

Method of introducing dlib to the Windows / Visual Studio environment

http://stackoverflow.com/questions/32736149/how-to-load-jpeg-file-using-dlib-libarary

After attempting to run a example program downloaded from Here, I understand for working with jpeg files , I must add #define DLIB_JPEG_SUPPORT directive to the project. but before that It's necessary to download jpeg library and add it to the project. I did These steps:

1.Download jpegsr9a.zip from here and unzipped it.

2.Download WIN32.mak and paste it into the jpeg root folder

3.Open Developer Command Prompt from visual studio 2013 tools

4.In command prompt type : nmake -f makefile.vc setup-v10

5.After these steps jpeg.sln created ,the note is when I open jpeg.sln in VS2013 the message come:

maybe base of the problem start from here , I don't know

6.Build the jpeg.sln with the proper configuration (I built it many times with different configurations, recently I built it using this .) at the end of building the error came :"unable to start jpeg.lib" but in release folder or debug folder (depend on configuration) jpeg.lib was created

7. open main project which is using DLIB for detecting face,I added jpeg root folder to Additonal Include Directory and jepegroot/release to Additional Libarary Directories ,then change the UseLibrary dependencies to "yes" and I also added jpeg.lib to the dependecies.

during building the project errors come:

This is the source which I trying to build and run

//#define HAVE_BOOLEAN
#define DLIB_JPEG_SUPPORT
#include <dlib/image_processing/frontal_face_detector.h>
#include <dlib/image_processing/render_face_detections.h>
#include <dlib/image_processing.h>
#include<dlib/image_transforms.h>
#include <dlib/gui_widgets.h>
#include <dlib/image_io.h>
#include <iostream>
//
using namespace dlib;
using namespace std;

// ----------------------------------------------------------------------------------------

int main(int argc, char** argv)
{
    try
    {
        // This example takes in a shape model file and then a list of images to
        // process.  We will take these filenames in as command line arguments.
        // Dlib comes with example images in the examples/faces folder so give
        // those as arguments to this program.
        if (argc == 1)
        {
            cout << "Call this program like this:" << endl;
            cout << "./face_landmark_detection_ex shape_predictor_68_face_landmarks.dat faces/*.jpg" << endl;
            cout << "\nYou can get the shape_predictor_68_face_landmarks.dat file from:\n";
            cout << "http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2" << endl;
            return 0;
        }

        // We need a face detector.  We will use this to get bounding boxes for
        // each face in an image.
        frontal_face_detector detector = get_frontal_face_detector();
        // And we also need a shape_predictor.  This is the tool that will predict face
        // landmark positions given an image and face bounding box.  Here we are just
        // loading the model from the shape_predictor_68_face_landmarks.dat file you gave
        // as a command line argument.
        shape_predictor sp;
         deserialize(argv[1])>>sp;


        image_window win, win_faces;
        // Loop over all the images provided on the command line.
        for (int i = 2; i < argc; ++i)
        {
            cout << "processing image " << argv[i] << endl;
            array2d<rgb_pixel> img;
            load_image(img, argv[i]);
            // Make the image larger so we can detect small faces.
            pyramid_up(img);

            // Now tell the face detector to give us a list of bounding boxes
            // around all the faces in the image.
            std::vector<rectangle> dets = detector(img);
            cout << "Number of faces detected: " << dets.size() << endl;

            // Now we will go ask the shape_predictor to tell us the pose of
            // each face we detected.
            std::vector<full_object_detection> shapes;
            for (unsigned long j = 0; j < dets.size(); ++j)
            {
                full_object_detection shape = sp(img, dets[j]);
                cout << "number of parts: " << shape.num_parts() << endl;
                cout << "pixel position of first part:  " << shape.part(0) << endl;
                cout << "pixel position of second part: " << shape.part(1) << endl;
                // You get the idea, you can get all the face part locations if
                // you want them.  Here we just store them in shapes so we can
                // put them on the screen.
                shapes.push_back(shape);
            }

            // Now let's view our face poses on the screen.
            win.clear_overlay();
            win.set_image(img);
            win.add_overlay(render_face_detections(shapes));

            // We can also extract copies of each face that are cropped, rotated upright,
            // and scaled to a standard size as shown here:
            dlib::array<array2d<rgb_pixel> > face_chips;
            extract_image_chips(img, get_face_chip_details(shapes), face_chips);
            win_faces.set_image(tile_images(face_chips));

            cout << "Hit enter to process the next image..." << endl;
            cin.get();
        }
    }
    catch (exception& e)
    {
        cout << "\nexception thrown!" << endl;
        cout << e.what() << endl;
    }
}

// ----------------------------------------------------------------------------------------

I can choose other alternatives but I spend too much time to reach here , I want to know How I can solve this problem and load jpeg file when using DLIB

I also read these links:

Compiling libjpeg

http://www.dahlsys.com/misc/compiling_ijg_libjpeg/index.html

dlib load jpeg files

http://sourceforge.net/p/dclib/discussion/442518/thread/8a0d42dc/

=======================================================================================

I solved my problem by below instruction, please follow it.

- Add include directory in VC++

- Include source.cpp

- Add add files in dlib/external/libjpeg to project

- Define in Preprocessor

-- You don't need to use any additional library.

http://gyeongju.tistory.com/entry/DLIB-C-Library

DLib 라이브러리 사용법

1. http://dlib.net/ 홈페이지 접속하여 라이브러리 다운로드

2. CMake 다운로드 https://cmake.org/download/ 

   Binary distributions: cmake-3.6.0-rc1-win64-x64.zip / cmake-3.6.0-rc1-win32-x86.zip

   컴퓨터 사양에 맞게 다운로드

   나의 경우 64비트 운영체제 이지만, 라이브러리 사용하는데 64, 32 둘다 필요한 경우가 있기 때문에

   둘다 다운받음

3. 현재 라이르러리 버전 "dlib-18.18"

   - zip 압축을 풀고, 빌드를 위해 폴더를 생성해 준다.

   - 나의 경우, dlib-18.18 → build       // 32비트

                   dlib-18.18 → build_64   // 64비트

   - CMake에서 다운받은 파일 압축풀고..

   - 64, 32 버전 둘다 bin 폴더에 들어가면 cmake-gui.exe 파일 있음 실행 고고

   - 윈도우 32비트 버전부터 고고

   - source code 에 내 dlib-18.18 폴더에 보면 examples 폴더가 있다. 선택해준다.

   - binaries : 3번에서 만들어준 빌드 폴더 선택한다. 지금 32비트버전이니 헷갈리지 않게 알아서 잘...

   - Configure 버튼을 한번 눌러준다.

   - 위 창이 뜬다. 현재 사용하고 있는 Visual studio 버전을 선택한다.

     그리고, visual studio 환경, 즉 32비트 개발환경인지 64비트 개발환경인지에 따라서 잘 선택해주면 된다.

     (목록에 64비트는 따로 있다.)

   - 나는 Visual studio 2010, 32비트니까 Visual Studio 10 2010 선택

   - 그다음 Finish 버튼 클릭

   - 아래와 같거나 비슷한 화면이 나온다.

   - 빨간색 부분에 필요한 부분이나 옵션 추가 선택한다.

   - Dlib 라이브러리를 사용한다면, 대부분 OpenCV 라이브러리는 사용하고 있는 분들이 많을 것 같다.

   - 없다면 OpenCV 다운로드!! http://gyeongju.tistory.com/entry/Visual-Studio-2010-OpenCV-243-설치

   - OpenCV_DIR 에 사용하고 있는 OpenCV 경로 선택해준다.

   * 체크박스 확인

   - DLIB_JPEG_SUPPORT

   - DLIB_PNG_SUPPORT

   - USE_AVX_INSTRUCTIONS

   - USE_SSE2_INSTRUCTIONS

   - USE_SSE4_INSTRUCTIONS

위에 USE_XXX_INSTRUCTIONS 는 실행속도와 관련있는 것으로 보이는데, (대충 봐서 기억안남)

라이브러리 중에서 위 세개 다 지원해야 돌아가는 함수도 있다. 그래서 난 다 선택해줌

   - 그리고  "Configure" 두번 클릭!! (빨간색 안나오면 된다)

   - 그 다음 "Generate" 클릭!!

     아래 창에 "Generating done" 뜨면 끝

     여기까지 CMake의 역할은 끝남.

  - C:\dlib-18.18\build 아까 내가 생성해준 build 폴더에 들어가면 뭐가 잔뜩 생성되어 있다.

  - C:\dlib-18.18\build\dlib_build 폴더에 들어가면 dlib.sln 파일이 있다. 클릭해서 비주얼스튜디오 열자

  - 빌드하자! Visual Studio → 빌드 → 구성관리자

  - 아래와 같이 세개 다 체크하고 빌드 시작. (시간이 좀 걸린다.)

 "Debug / Release" 둘다 빌드하면 "dlib_build" 폴더 안에 Debug, Release 폴더가 생기고,

    그 안에 보면 dlib.lib 파일이 생성되어 있다.

여기까지, DLib 를 사용하기 위해 Cmake를 이용한 컴파일 끝!!

64비트 버전은 62비트 CMake를 실행하여 진행하면 된다.

(선택해야 하는 부분이 있으면 64버전으로 다 선택하면 되고, 과정은 똑같다.)

그 다음은 실제 라이브러리를 사용하기 위한 연동방법에 대해 정리해야 겠다.