c++ - Is it possible to load/read shape_predictor_68_face_landmarks.dat at compile time? -

i trying build c++ application in visual studio using dlib's face_landmark_detection_ex.cpp. build application run command promt , trained model , image file passed arguments.

face_landmark_detection_ex.exe shape_predictor_68_face_landmarks.dat image.jpg 

this shape_predictor_68_face_landmarks.dat trained model 68 landmarks perform detection on input image , needs load @ run-time every time perform detection. trying following things.

• load shape_predictor_68_face_landmarks.dat @ building application or compile time.
• read shape_predictor_68_face_landmarks.dat inside code every time application strarts execution, not take more amount of memory.

is there way pack file inside application take less physical memory run.

update:

how can store shape_predictor_68_face_landmarks.dat file in static buffer every time shape_predictor can read buffer.

yes, possible, depends on visual studio , not cross-platform

1. you should create resource file , include hape_predictor_68_face_landmarks.dat project. see https://msdn.microsoft.com/ru-ru/library/7zxb70x7.aspx details. make compiler put file exe/dll

2. open resoure @ runtime , memory pointer https://msdn.microsoft.com/en-us/library/windows/desktop/ee719660(v=vs.85).aspx

3. create memory stream (std::istream) pointer.

   4. deserialize stream dlib::deserialize

here minimal example, without resource reading:

#include <string> #include <iostream> #include <dlib/image_processing/shape_predictor.h>   struct membuf : std::streambuf {     membuf(char const* base, size_t size) {         char* p(const_cast<char*>(base));         this->setg(p, p, p + size);     } }; struct imemstream : virtual membuf, std::istream {     imemstream(char const* base, size_t size)             : membuf(base, size)             , std::istream(static_cast<std::streambuf*>(this)) {     } }; using namespace dlib; //its important use namespace dlib deserialize work correctly using namespace std; int main(int argc, const char* argv[]) {     const char* file_name = "shape_predictor_68_face_landmarks.dat";     ifstream fs(file_name, ios::binary | ios::ate);     streamsize size = fs.tellg();     fs.seekg(0, ios::beg);     std::vector<char> buffer(size);     if (fs.read(buffer.data(), size))     {         cout << "successfully read " << size << " bytes " << file_name << " buffer" << endl;         imemstream stream(&buffer.front(), size); // here loading memory buffer. can change line use pointer resource         shape_predictor sp;         deserialize(sp, stream);         cout << "deserialized shape_predictor" << endl;     }     else cout << "failed read " << file_name << " buffer" << endl;     return 0; } 

and memory usage.

first of should know shape_predictor::operator() const, , documentation says safe use 1 shape_predictor different threads.

so, can create 1 shape_predictor @ start of program , use many times, different threads

next, putting shape predictor inside resource make loaded ram when program starts, deserializing resource make copy of memory, , lead ram usage overhead. if need minimal possible ram usage - should load file

and last question - how initialize compiler. there no ready-to-use solution it, can use code shape_predictor.h/deserialize function , load manually. think, bad solution, because not less ram usage compared loading file

so recommendation load 1 shape_predictor file , use globally threads