/octuplet-loss

Octuplet Loss: Make Your Face Recognition Model Robust to Image Resolution

Primary LanguagePythonMIT LicenseMIT

Octuplet Loss - Make Face Recognition Robust Against Image Resolution

Code style: black License Last Commit

Here, we release our code utilized in the following paper:

Loss Visualization

🏆 Performance (Accuracy [%])

Model LFW XQLFW
ArcFace 99.55 93.27
MagFace 99.63 92.92
FaceTransformer 99.73 95.12

💻 Code

We provide the code of our proposed octuplet loss for tf_octuplet_loss.py (Tensorflow 2) and pt_octuplet_loss.py (PyTorch).

🥣 Requirements

Python 3.8

📖 Cite

If you use our code please consider citing:

@misc{https://doi.org/10.48550/arxiv.2207.06726,
  doi = {10.48550/ARXIV.2207.06726},
  url = {https://arxiv.org/abs/2207.06726},
  author = {Knoche, Martin and Elkadeem, Mohamed and Hörmann, Stefan and Rigoll, Gerhard},
  keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences},
  title = {Octuplet Loss: Make Face Recognition Robust to Image Resolution},
  publisher = {arXiv},
  year = {2022},
  copyright = {arXiv.org perpetual, non-exclusive license}
}

✉️ Contact

For any inquiries, please open an issue on GitHub or send an E-Mail to: Martin.Knoche@tum.de