openface, dlib, opencv

install dlib

make sure your cmake version > 2.8

cmake --version

upgrade

sudo add-apt-repository ppa:george-edison55/cmake-3.x
sudo apt-get update
sudo apt-get install cmake

we need boostpython

sudo apt-get install libboost-all-dev

install dlib, it work in virtualenv too

pip install dlib

install torch

before you install it, think one second

cuda support?

you must install cuda toolkit before install torch, that will auto config for that

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

with follow lib

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

openface

Face recognition with deep neural networks

git clone and python setup.py install

edit site-packages/openface/torch_neural_net.py

line 130, 131 about the string and bytes convert

refer

self.p.stdin.write((imgPath + "\n").encode(encoding='utf-8'))
output = self.p.stdout.readline().decode(encoding='utf-8')

monitor gpu

here is monitor nividia gpu usage

watch -n 0.5 nvidia-smi