Example of small shared library, an dlib wrapper for face tracking and SVM based emotion classification.
Dlib is powerfull machine learning library and an excelent example of an academic c++ coding style. High level of abstraction in the library and abundant templating - leаd to very long compilation time :( Compilation of any simplest app, risked to include some dlib headers, become endless.
So, isolation of some dlib based functionality in sepparate shared library, sounds good.
- cmake
- dlib
- SDL for clients
- zmq for zmq client/server
- edit CMakeList.txt, set your dlib location path
cd build; cmake ..; make
./face_track_test
- face_tracker.so - shared library
- face_track_test - library test
- face_track_[server/client] - shared memory communication
- face_track_[server/client]_zmq - zmq based communication