---
license: mit
language:
- en
---

# Hyp-OC Model Card

<div align="center">

[**Project Page**](https://kartik-3004.github.io/hyp-oc/) **|** [**Paper (ArXiv)**](https://arxiv.org/pdf/2404.14406.pdf) **|** [**Code**](https://github.com/Kartik-3004/hyp-oc)


</div>

## Introduction

Hyp-OC, is the first work exploring hyperbolic embeddings for one-class face anti-spoofing (OC-FAS). 
We show that using hyperbolic space helps learn a better decision boundary than the Euclidean counterpart, 
boosting one-class face anti-spoofing performance.

<div  align="center">
<img src='assets/visual_abstract.png' height="50%" width="50%">
</div>

## Training Framework
<div  align="center">
<img src='assets/framework.png'>
</div>

Overview of the proposed pipeline: Hyp-OC. The encoder extracts facial features which are used to estimate the mean of Gaussian 
distribution utilized to sample pseudo-negative points. The real features and pseudo-negative features are then concatenated 
and passed to FCNN for dimensionality reduction. The low-dimension features are mapped to Poincaré Ball using *exponential map*. 
The training objective is to minimize the summation of the proposed loss functions Hyp-PC} and Hyp-CE. The result is a separating 
*gyroplane* beneficial for one-class face anti-spoofing.



## Usage

The pre-trained weights can be downloaded directly from this repository or using python:
```python
from huggingface_hub import hf_hub_download

hf_hub_download(repo_id="kartiknarayan/hyp-oc", filename="pretrained_weights/vgg_face_dag.pth", local_dir="./")
```

## Citation
```bibtex
@article{narayan2024hyp,
  title={Hyp-OC: Hyperbolic One Class Classification for Face Anti-Spoofing},
  author={Narayan, Kartik and Patel, Vishal M},
  journal={arXiv preprint arXiv:2404.14406},
  year={2024}
}
```

Please check our [GitHub repository](https://github.com/Kartik-3004/hyp-oc) for complete instructions.