1. apt-get install libopencv-dev (caution: installs OpenCV version 2.3.x and it may not contain the "-lopencv_nonfree" lib. So compile it yourself.)
2. apt-get install qt4-dev-tools
3. Download QCustomPlot (http://www.qcustomplot.com/). Copy qcustomplot.cpp and qcustomplot.h to your lr-facetags directory.

You can check the available libs with: pkg-config --libs opencv

1. export QT_SELECT=4
2. qmake
3. make
4. ./lr-facetags

For the datatool usage see: Datasettools2

1. Select a image folder with the "Open Folder" button or by typing the path into box beside the button

• Subfolders are not considered
• The image folder path musst always be correct

2. Face Detection

• Select the detection classifier in the drop down box
• Press "Detect Faces"

3. Face Recognition

• Check that you detected Faces at least once
• Select the recognition algorithm in the drop down box
• Press "Recognize Faces"

4. Benchmarking

• Check if the ground truth exists, if not create one with the datatool
• For each detection run a subfolder with the timestamp will be created in: ImageFolder/.metaface/TIMESTAMP and a target will be added to the benchmarking selection
• By pressing "Laod Single Run" or "Load All Runs" you can previous runs for benchmarking
• Press "Compare Detection" to start the face detection benchmarking
• Press "Compare Recognition" to start the face recognition benchmarking