INSTALLATION GUIDE

Dependencies

Download libsvm from LINK. Move it to the root folder of the project and rename it libsvm. To compile it, go in Matlab to libsvm/matlab and run the function make.m.
Download matconvnet from LINK. Move it to the root folder of the project and rename it matconvnet. To compile it, go in Matlab to matconvnet/matlab and run vl_compilenn.
Download VGG Face Matconvnet from LINK. Move it to the root folder of the project and rename it vgg_face_matconvnet. If the mex-files for your system are not present, copy the mex-files for vl_nnconv and vl_nnpool from matconvnet/matlab/simplenn to vgg_face_matconvnet/+lib/+face_feats/@convNet/ .
Download VLFeat binary package (that already contains the compiled mex files) from LINK.
Download DLib from LINK. Move it to the root folder of the project and rename it dlib. First copy face_detector/face_detector.cpp to dlib/examples. Then to compile it, go to dlib/examples/ with a terminal and run:
```
 mkdir build
 cd build
 cmake ..
 cmake --build .
```
Download the JDBC driver from LINK to be able to use Mysql with Matlab. To install it, follow the instruction at LINK and at LINK.
Other dependencies:
- Stack LAMP.
- MySQLdb and bs4 (BeautifulSoup) library for python.

Setup

Go to root_of_project/ImagesCollector/conf/, make a copy of proto.conf and call it config.conf. Then fill the parameters inside the configuration file.
Insert a list of identities (identities.mat or identities.txt) in DATA_PATH/data/.

Web App

Copy the application to /var/www/.
Configure database.php with the database called DATABASE_DATASET in config.conf.
Make a symbolic link to the folder where the cropped images are, the same path used for DATA_PATH in config.conf:
```
ln DATA_PATH/img_crop/ /var/www/name_of_the_web_app/media
```