FaceRecognizer

Face Recognition - a machine learning way

:Author: Sourav Senapati

What is about?

Facial Recognition sucks. But its getting better. This isa trivial toy project to learn how good Facial Recognition is.

Currently, the work is on face recognition but soon, I will be updating the code base for face, gender detection and cool face tracking.

Compile

This is a CMake project. CMake takes a CMakeList file, and outputs to a platform-specific build format. If in Linux system, create a build directory, preferably parallel to source directory (src) and go to the buld directory. To achieve that, open a terminal and type the following command:

$ mkdir build_x86-64
$ cd build_x86-64

Then run cmake and point to the source directory. Type the followiing command:

$ cmake ../src

And finally run make in the build directory to generate the executables, the executables are found under application folder. Type with the following command:

$ make

ECLIPSE EDITOR : If Eclipse is your favorite editor, then a small script is being provided for making an Eclipse build project. Run the following command in the source folder while providing the arguments:

$ ./mk-eclipse-build-project.sh SRC_DIR BUILD_DIR BUILD_TYPE

Run

To run the application, open a terminal, go to the build directory and type the follwing:

$ ./application/FaceRecognizerApp

The results are shown on the terminal console.

Document generation

Support for Doxygen document generation system has been supplied with the project. A CMake file is provided that generates required config file in the build directory. After successfully building the project, open a terminal, go to the build directory and type:

$ doxygen Doxyfile

This generates Doxygen docuementation inside doxygen folder for the current project. The document generation is both in HTML and Latex. In case the doxygen doc is not required, please comment out the follwing line in CMakeLists.txt file unders source folder:

include(cmake/doxygen.cmake)