---
license: mit
language:
- en
library_name: diffusers
---

# Arc2Face Model Card

<div align="center">

[**Project Page**](https://arc2face.github.io/) **|** [**Paper (ArXiv)**](https://arxiv.org/abs/2403.11641) **|** [**Code**](https://github.com/foivospar/Arc2Face) **|** [🤗 **Gradio demo**](https://huggingface.co./spaces/FoivosPar/Arc2Face)



</div>

## Introduction

Arc2Face is an ID-conditioned face model, that can generate diverse, ID-consistent photos of a person given only its ArcFace ID-embedding.
It is trained on a restored version of the WebFace42M face recognition database, and is further fine-tuned on FFHQ and CelebA-HQ.

<div  align="center">
<img src='assets/samples_short.jpg'>
</div>

## Model Details

It consists of 2 components:
- encoder, a finetuned CLIP ViT-L/14 model
- arc2face, a finetuned UNet model

both of which are fine-tuned from [runwayml/stable-diffusion-v1-5](https://huggingface.co./runwayml/stable-diffusion-v1-5).
The encoder is tailored for projecting ID-embeddings to the CLIP latent space.
Arc2Face adapts the pre-trained backbone to the task of ID-to-face generation, conditioned solely on ID vectors.

## ControlNet (pose)

We also provide a ControlNet model trained on top of Arc2Face for pose control.

<div  align="center">
<img src='assets/controlnet_short.jpg'>
</div>

## Usage

The models can be downloaded directly from this repository or using python:
```python
from huggingface_hub import hf_hub_download

hf_hub_download(repo_id="FoivosPar/Arc2Face", filename="arc2face/config.json", local_dir="./models")
hf_hub_download(repo_id="FoivosPar/Arc2Face", filename="arc2face/diffusion_pytorch_model.safetensors", local_dir="./models")
hf_hub_download(repo_id="FoivosPar/Arc2Face", filename="encoder/config.json", local_dir="./models")
hf_hub_download(repo_id="FoivosPar/Arc2Face", filename="encoder/pytorch_model.bin", local_dir="./models")
```

Please check our [GitHub repository](https://github.com/foivospar/Arc2Face) for complete inference instructions.

## Limitations and Bias

- Only one person per image can be generated.
- Poses are constrained to the frontal hemisphere, similar to FFHQ images.
- The model may reflect the biases of the training data or the ID encoder.

## Citation


**BibTeX:**

```bibtex
@misc{paraperas2024arc2face,
      title={Arc2Face: A Foundation Model of Human Faces}, 
      author={Foivos Paraperas Papantoniou and Alexandros Lattas and Stylianos Moschoglou and Jiankang Deng and Bernhard Kainz and Stefanos Zafeiriou},
      year={2024},
      eprint={2403.11641},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
```