Skip to content

Latest commit

 

History

History

evaluations

readme.md

FDFR and ISM

Set up env for FDFR and ISM. Firstly, we need to cd evalutions and install the following library:

cd deepface
pip install -e .

cd retinaface
pip install -e .

Note: You should follow the tensorflow instruction from https://www.tensorflow.org/install/pip to accelerate evaluation process (without tensorflow, the eval code will run on cpu which is slow)

To compute FDFR and ISM. If you get block by proxy, manually download weight of DeepFace from https://github.com/serengil/deepface_models/releases/download/v1.0/arcface_weights.h5 and download weight from https://github.com/serengil/deepface_models/releases/download/v1.0/retinaface.h5, place it in folder "{home}/.deepface/weights/". We run the following command to compute FDFR and ISM soore: python evaluations/ism_fdfr.py --data_dir <path_to__perturb_image_dir> --emb_dirs <paths_to_id_emb_file>

SER-FIQ

  1. cd FaceImageQuality
  2. download the model files and place them in the insightface/model and install environment by command pip install -r requirements.txt (Note: should use numpy==1.22.0)
  3. Run python evaluations/ser_fiq.py --prompt_path <path_to_perturb_images_of_prompts> --gpu 0 where gpu is the cuda device id

Brisque

  1. pip install brisque
  2. Run python evaluations/brisque.py --prompt_path <path_to_perturb_images_of_prompts>