FaceDetector

Install

apt dependencies

sudo apt-get install python-pip cmake python-opencv python-tk python-imaging python-imaging-tk libjpeg-dev libfreetype6 libfreetype6-dev zlib1g-dev libboost-all-dev

python dependencies

sudo pip2 install -r requirements.txt

install torch

git clone https://github.com/torch/distro.git ~/torch --recursive
cd ~/torch; bash install-deps;
./install.sh

install openface dependencies

sudo apt-get install luarocks
for NAME in dpnn nn optim optnet csvigo cutorch cunn fblualib torchx tds; do luarocks install $NAME; done

install openface

git clone https://github.com/cmusatyalab/openface/ ~/openface --recursive
cd ~/openface
sudo python2 setup.py install
bash models/get-models.sh

breath sigh of relief, well done!

Use

Install

Run FaceDetector.py

Enable or disable different filters in the menu

TODO

CV2

Re-implement training as another layer

Dlib

Implement training in Dlib

Train Dlib with other objects

• Eyes
• Hands
• Geometric shapes
• Other simple objects

Data sets

Hundreds of datasets http://homepages.inf.ed.ac.uk/rbf/CVonline/Imagedbase.htm#face