Skip to content

Research project for testing the techniques of processing the human face images via the real-time video streaming

Notifications You must be signed in to change notification settings

oradzhabov/FakeFace

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FakeFace

Research project for testing the techniques of processing the human face images via the real-time video streaming.

It uses DirectShow filters for grabing video frames from real camera device, processing them and passing to the virtual camera.

During processing of frames it uses dlib or parts of OpenFace to putting any human face from image to corresponded face on the frame.

Alt text

Minimum Requirements

  • Windows OS (x86 & x64)
  • Microsoft Visual Studio 2010

Dependencies

OpenCV

dlib

  • web: http://dlib.net/
  • github: https://github.com/davisking/dlib
  • notes:
    1. dlib should not be just compiled but installed. Hence in my case, when I point dlib to install in local folder, I have next path for cmake entry dlib_DIR D:/DLIB/dlib-18.18/build/msvc2010.x32/install/lib/cmake/dlib
    2. I've used dlib 18.18

boost

Microsoft Windows SDK

Build

Use only release target

Prepare

Register new virtual camera

  • Run FAR or CMD with Administrator permissions
  • Point to folder where file VirtualCamDevice.dll has been built
  • During building there are several files has been copied to the same folder
  • Run file install.bat
  • If no errors occured during registration system will inform about it twice (actually 2 files will be registered)

Unregister virtual camera

  • If you do not want to continue using the virtual camera, run file uninstall.bat from the folder where file VirtualCamDevice.dll has been built
  • Ensure that no applications (GraphStudio, Skype, Hangout or else) do not use virtual camera. Otherwise it will not uninstall it

Run

  • Ensure that no application uses your real web-camera device. Otherwise it could not be stated as imput source for FakeFace
  • Run file VirtualCameraManager.exe
  • After several seconds FakeFace manager window will apear
  • From this moment you can use FakeFace

Using Cases

GraphStudio

  • For testing purposes I use Graphstudio (http://blog.monogram.sk/janos/2009/06/14/monogram-graphstudio-0320/)
  • Notice that if you build FakeFace for x64 platform, use GraphStudio64 instead of GraphStudio
  • Download and run it. Open project "./data/MONOGRAM GraphStudio/FakeFace Virtual Cam.grf"
  • This project connect virtual camera (registered before) and renderer
  • Start the graph
  • If you did not start VirtualCamManager, renderer will show noisy pixels
  • If VirtualCamManager has been started before, you will see your web-camera stream where your face will be switched to face of Vasyl Lomachenko - default fake face
  • Pick to the FakeFace Manager window and press "S"-button on the keyboard
  • OpenFile dialog will appear. Choose some file with human face photo
  • After moment your face on the GraphStudio renderer will changed to that which imaged on the selected file

Skype

  • Make sure that the destination platform of the installed version of Skype corresponds to the build destination platform of FakeFace
  • If you wish to use FakeFace with Skype, you need to register virtual camera when Skype is turned off. Remember that virtual camera should be registered when no one application use your real camera device
  • After virtual camera has been registered, VirtualCamManager has been started and Skype has been stared pick Skype menu: Tools >> Options >> Video settings
  • Your default camera will not be available
  • On the panel "Video settings" select webcam: "FakeFace Virtual Cam"
  • Enjoy

Tips

  • Avoid back lighting. Set light direction frontal to face
  • Use clothes that cover your neck. Color of your face shold not interfere with color of neck as background color

About

Research project for testing the techniques of processing the human face images via the real-time video streaming

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages