Susceptibility to Image Resolution in Face Recognition and Training Strategies

Evaluation Protocols presented in the paper can be found here in this repository.

The .csv-files contain the downsampling resolution for each consecutive image in the corresponding test dataset.

📚 Citations

If you want to use our provided evaluation protocols please cite our paper: Image Resolution Susceptibility of Face Recognition Models

@article{Knoche_Hörmann_Rigoll_2022, 
place={Wadern, Germany}, 
title={Susceptibility to Image Resolution in Face Recognition and Training Strategies to Enhance Robustness}, 
volume={8}, 
url={https://ojs.dagstuhl.de/index.php/lites/article/view/lites-v008-i001-a001}, 
DOI={10.4230/LITES.8.1.1}, number={1}, journal={Leibniz Transactions on Embedded Systems}, 
author={Knoche, Martin and Hörmann, Stefan and Rigoll, Gerhard}, 
year={2022}, 
month={Nov.}, 
pages={01:1–01:20}
}

Martlgap/btm-stm

Susceptibility to Image Resolution in Face Recognition and Training Strategies

📚 Citations