---
license: mit
pretty_name: "Serpent"
viewer: false
---
<h1 align="center">Serpent: Scalable and Efficient Image Restoration via Multi-scale Structured State Space Models</h1>
<!-- <h3 align="center">Mohammad Shahab Sepehri, Zalan Fabian, Maryam Soltanolkotabi, Mahdi Soltanolkotabi</h3> -->

<p align="center">
  <a href="https://scholar.google.com/citations?user=j2scUKoAAAAJ&hl=en">Mohammad Shahab Sepehri</a> 
  <a href="https://scholar.google.com/citations?user=5EKjsXQAAAAJ&hl=en">Zalan Fabian</a> 
  <a href="https://scholar.google.com/citations?user=narJyMAAAAAJ&hl=en">Mahdi Soltanolkotabi</a> 
</p>

<p align="center">
  | <a href="https://arxiv.org/abs/2403.17902">Paper</a> 
  |
  <a href="https://github.com/AIF4S/Serpent">Github Repository</a> 
  |
</p>


[![License](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/license/MIT)

<p align="justify" > 
<strong>Serpent</strong> is a novel architecture for efficient image restoration that leverages state space models capable of modeling intricate long-range dependencies in high-resolution images with a favorable linear scaling in input dimension.
  <br />
You can download our pretrained models from this repository
</p>

## Usage
You can find our code and instructions for using our pre-trained models on [our Github repository](https://github.com/AIF4S/Serpent).