FDDB-Evaluation

Python Evaluation Code for FDDB Face Dataset

Usage

before evaluating

python3 setup.py build_ext --inplace

evaluating

GroungTruth: under the ground_truth

Notice: The prediction directory should be like widerface dataset. I generate the fake prediction result under pred_sample for you reference.

python3 evaluate.py -p <your prediction dir> -g <groud truth dir>

Acknowledgements

some code borrowed from Widerface-Evaluation