- OpenCV
- CMake
- Python
Build
run the following command:
git clone https://github.com/GreenKing/fddb-evaluation
# ${fddb-evaluation}
cd fddb-evaluation/fddb-evaluation
mkdir build
cd build
cmake .. -G"Visual Studio 12 2013 Win64"
Open fddb-eval.sln
file to compile fddbEval.exe
. It should be noticed that it doesn't work with release
mode in my computer. So you should be very careful if you use the release
mode.
After building the runable file,
cd ${fddb-evaluation}
cp fddb-evaluation/DEBUG/fddbEval.exe bin/
fddbEval usage
./fddbEval [OPTIONS]
-h : print usage
-a fileName : file with face annotations (default: ellipseList.txt)
-d fileName : file with detections (default: faceList.txt)
-f format : representation of faces in the detection file (default: 0)
[ 0 (rectangle), 1 (ellipse), or 2 (pixels) ]
-i dirName : directory where the original images are stored
(default: ~/scratch/Data/facesInTheWild/)
-l fileName : file with list of images to be evaluated (default: temp.txt)
-r fileName : prefix for files to store the ROC curves (default: temp)
-s : display the matching detection-annotation pairs.
-z imageFormat : image format used for reading images for the annotation set
(default: .jpg )
for example
fddbEval -a ./FDDB-folds/FDDB-fold-ellipseList.txt -d ./fddb.txt -i path/to/images/dir -l ./FDDB-folds/FDDB-folds.txt -r ./facedet
Evaluation
cd ${fddb-evaluation}
python runEvaluate.py eval 1 facedet