/Plastic-Surgery-Recognition

An application that diagnoses whether the nose is aesthetic in real time using tensorflow, keras and dlib.

Primary LanguagePythonMIT LicenseMIT

Plastic Surgery Recognition

An application that diagnoses whether the nose is aesthetic in real time using tensorflow, keras and dlib.

XhW

Setup This Project

Install DLIB & OpenCV

  • You can install the dlib library by typing conda install -c conda-forge dlib in your terminal. Anaconda must be installed.
  • The easiest way to install opencv is to download it from PyPI. It's going to install the library itself and its prerequisites as well. You can install the opencv library by typing pip install opencv-python in your terminal.
  • Also install tensorflow and keras too
  • And then you can run the project with model.py first. It will create the model.h5. Then you can run main.py. Make sure you have "shape_predictor_68_face_landmarks.dat" file in your project location.
  • Download shape_predictor_68_face_landmarks.dat I could not add it to this project because it is larger than 25mb

How this is working?

We created a model with images from our dataset. (resnet50) Used dlib and dlib's face alignment also opencv. Opencv gets realtime images from our camera and we diagnose each fragment with our model.