"Openface" _ Deep Learning

Openface:

http://cmusatyalab.github.io/openface/

One, what is openface.

Openface is a open source face recognition system based on deep neural network. The system is based on Google's article Facenet:a Unified embedding for face recognition and clustering. Openface is a leading Brandon Amos at Carnegie Mellon University.

Official address: http://cmusatyalab.github.io/openface/

Code: Https://github.com/cmusatyalab/openface Second, the OPENFACE environment constructs

System: Ubuntu 14.04 64-bit desktop operating system

Reference: HTTP://CMUSATYALAB.GITHUB.IO/OPENFACE/SETUP/1, Ubuntu switch root user

If you want to use ordinary users, please test yourself.

：

http://blog.csdn.net/itas109/article/details/50679251

2. Pre-Installation Preparation work

Install the necessary programs, you can use the following batch or one for installation.

  [cpp]  View plain  copy   #!/bin/sh   sudo apt-get install  build-essential -y   sudo apt-get install cmake -y   sudo  apt-get install curl -y   sudo apt-get install gfortran -y    sudo apt-get install git -y   sudo apt-get install  libatlas-dev -y   sudo apt-get install libavcodec-dev -y   sudo  apt-get install libavformat-dev -y   sudo apt-get install  libboost-all-dev -y   sudo apt-get install libgtk2.0-dev -y   sudo  apt-get install libjpeg-dev -y   sudo apt-get install  liblapack-dev -y   sudo apt-get install libswscale-dev -y   sudo  apt-get install pkg-config -y   sudo apt-get install python-dev -y   sudo apt-get  install python-pip -y   sudo apt-get install wget -y   sudo  apt-get install zip –y   [cpp]  View plain  copy   #!/bin/sh    sudo apt-get install build-essential -y   Sudo apt-get install  cmake -y   sudo apt-get install curl -y   Sudo apt-get  install gfortran -y   sudo apt-get install git -y   sudo  apt-get install libatlas-dev -y   sudo apt-get install  libavcodec-dev -y   sudo apt-get install libavformat-dev -y   sudo  apt-get install libboost-all-dev -y   sudo apt-get install  libgtk2.0-dev -y   sudo apt-get install libjpeg-dev -y   sudo apt-get  install liblapack-dev -y   sudo apt-get install libswscale-dev -y    sudo apt-get install pkg-config -y   sudo apt-get install  python-dev -y   sudo apt-get install python-pip -y   sudo  apt-get install wget -y   sudo apt-get install zip –y  

3. Install the necessary library [CPP] View plain copy pip2 install numpy scipy pandas pip2 install Scikit-learn scikit-image [CPP] View Plain copy pip2 install numpy scipy pandas pip2 install Scikit-learn scikit-image

Attention:

A. If one of the installation failures occurs, you can install one at a

B. Increase PIP installation speed

Can replace the PIP image to speed up download
Create./pip/pip.conf, enter the following (or other available mirrors): [CPP] View plain copy [Global] timeout = 6000 Index-url = http://pypi.douban.co m/simple [Install] use-mirrors = true mirrors = <a target=_blank href= "http://pypi.douban.com/" >http://pypi      .douban.com/</a> [CPP] View plain copy [Global] timeout = 6000 Index-url = http://pypi.douban.com/simple [Install] Use-mirrors = true mirrors = <a target=_blank href= "http://pypi.douban.com/" >http://pypi.douban.co M/</a>

C. Error: Sslerror:the read operation timed out

You can use the following instructions to add a delay [CPP] View plain copy pip2-install scikit-image--timeout [CPP] View plain copy Pip2-install Scikit-image--timeout 100

4. Installation Torch

A. Installation dependencies [CPP] View plain copy curl-shttps://raw.githubusercontent.com/torch/ezinstall/master/install-deps | bash–e [CPP] View plain copy curl-shttps://raw.githubusercontent.com/torch/ezinstall/master/install-deps | Bash–e

B. Install [CPP] View plain copy git clone https://github.com/torch/distro.git ~/torch--recursive cd ~/torch &&. /install.sh [CPP] View plain copy git clone https://github.com/torch/distro.git ~/torch--recursive CD ~/torch &am p;&./install.sh

C. Installation dependencies

Luarocks install $NAME, where $NAME is as listed below.

Dpnn

Nn

Csvigo

Cunn (using Cuda)

Fblualib (only for training DNN)

Torchx (only for training DNN)

command line, install as needed: [cpp]  view plain  copy   ~/torch/install/bin/luarocks install dpnn   ~/torch/install/bin/luarocks install nn   ~/torch/install/bin/luarocks install  optim   ~/torch/install/bin/luarocks install csvigo   ~/torch/install/bin/luarocks  install cunn   ~/torch/install/bin/luarocks install fblualib   ~/torch/ install/bin/luarocks install torchx   [cpp]  View plain  copy   ~/torch/install/ bin/luarocks install dpnn   ~/torch/install/bin/luarocks install nn   ~/ torch/install/bin/luarocks install optim   ~/torch/install/bin/luarocks install  csvigo   ~/torch/install/bin/luarocks install cunn   ~/torch/install/bin/luarocks  install fblualib   ~/TORCH/INSTALL/BIN/LUAROCKS&NBSP;INSTALL&NBSP;TORCHX&NBSP;&NBsp

D. Verify that the installation depends on success

Validate with th command

Attention:

A.gitclone Update Network old interrupt

Git submodule Update--init–recursive

or the Torch directory.

update.sh

Suggest using update.sh to solve

B. Error:

Cloning into ' EXTRA/LUAFFIFB ' ...

Remote:counting objects:918, done.

error:rpcfailed; result=56, HTTP code = 200| 0 bytes/s

Fatal:theremote End hung up unexpectedly

Fatal:earlyeof

Fatal:index-pack failed

Clone of ' HTTPS://GITHUB.COM/FACEBOOK/LUAFFIFB ' intosubmodule path ' EXTRA/LUAFFIFB ' failed

Resolved: [cpp]  view plain  copy   git config --global http.postbuffer 524288000