Evaluation

Running Evaluation

The script backend/face_recognition/evaluate.py evaluates the pipeline as well as the model:

--model-dir: Path to model (default: backend/face_recognition/results/models/facesecure_model.pth)

--pretrained-dir: Path to pretrained model (default: backend/face_recognition/results/models/pretrained.pt)

--eval-data: Path to evaluation data (default: backend/face_recognition/data/images/lfw_overall_eval_all)

E.g. run:

$ cd backend
$ python face_recognition/evaluate.py --model-dir <path to model> --pretrained-dir <path to pretrained model> --eval-data <path to evaluation data>

Please refer to Filestructure for the filestructure of the evaluation data.