laplab/facial-keypoints

Facial keypoints detection using CNN

Facial Keypoints Detection

Installation

Script is tested on Python 3.6.1. To install required libraries run:

pip install -r requirements.txt

Train data

Data is expected to be a set of face photos and CSV file where each row represents coordinates (xi, yi) of facial keypoints for image filename:

filename	x0	y0	x1	y1	...	x13	y13

Coordinates are indexed according to the following scheme:

Pretrained weights

You can use weights from 500 epochs training saved in weights/conv5_adam_epochs500.hdf5.
MSE error on images resized to (100, 100) shape is 5.2569.

Usage

Augmenting data

$ python app.py augment --help
Usage: app.py augment [OPTIONS] img coords dest

  Augments images stored in IMG folder with coordinates from COORDS csv file
  and saves result in grayscale to DEST folder

Training model

$ python app.py train --help  
Usage: app.py train [OPTIONS] img coords model

  Trains model on images from IMG folder with coordinates from COORDS csv
  file and saves trained model in hdf5 file MODEL

Predicting keypoints

$ python app.py predict --help
Usage: app.py predict [OPTIONS] img model coords

  Predicts facial keypoints coordinates for images from IMG folder using
  model from file MODEL and saves results in csv file COORDS

Results

These are some examples detected using pretrained weights: