## Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic Image Restoration In the Wild

> [[Paper](https://arxiv.org/abs/2401.13627)] &emsp; [[Project Page](http://supir.xpixel.group/)] &emsp; [Online Demo (Coming soon)] <br>
> Fanghua, Yu, [Jinjin Gu](https://www.jasongt.com/), Zheyuan Li, Jinfan Hu, Xiangtao Kong, [Xintao Wang](https://xinntao.github.io/), [Jingwen He](https://scholar.google.com.hk/citations?user=GUxrycUAAAAJ), [Yu Qiao](https://scholar.google.com.hk/citations?user=gFtI-8QAAAAJ), [Chao Dong](https://scholar.google.com.hk/citations?user=OSDCB0UAAAAJ) <br>
> Shenzhen Institute of Advanced Technology; Shanghai AI Laboratory; University of Sydney; The Hong Kong Polytechnic University; ARC Lab, Tencent PCG; The Chinese University of Hong Kong <br>


<p align="center">
  <img src="assets/teaser.png">
</p>

---
## 🔧 Dependencies and Installation


1. Clone repo
    ```bash
    git clone https://github.com/Fanghua-Yu/SUPIR.git
    cd SUPIR
    ```

2. Install dependent packages
    ```bash
    conda create -n SUPIR python=3.8 -y
    conda activate SUPIR
    pip install --upgrade pip
    pip install -r requirements.txt
    ```

3. Download Checkpoints
#### Dependent Models
* [SDXL CLIP Encoder-1](https://huggingface.co/openai/clip-vit-large-patch14)
* [SDXL CLIP Encoder-2](https://huggingface.co/laion/CLIP-ViT-bigG-14-laion2B-39B-b160k)
* [SDXL base 1.0_0.9vae](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_base_1.0_0.9vae.safetensors)
* [LLaVA CLIP](https://huggingface.co/openai/clip-vit-large-patch14-336)
* [LLaVA v1.5 13B](https://huggingface.co/liuhaotian/llava-v1.5-13b)


#### Models we provided:
* `SUPIR-v0Q`: (Coming Soon) Google Drive, Baidu Netdisk
    
    Default training settings with paper. High generalization and high image quality in most cases.

* `SUPIR-v0F`: (Coming Soon) Google Drive, Baidu Netdisk

    Training with light degradation settings. Stage1 encoder of `SUPIR-v0F` remains more details when facing light degradations.

4. Edit Custom Path for Checkpoints
    ```
    * [CKPT_PTH.py] --> LLAVA_CLIP_PATH, LLAVA_MODEL_PATH, SDXL_CLIP1_PATH, SDXL_CLIP2_CACHE_DIR 
    * [options/SUPIR_v0.yaml] --> SDXL_CKPT, SUPIR_CKPT_Q, SUPIR_CKPT_F
    ```
---

## ⚡ Quick Inference


### Usage of SUPIR

```console
Usage: 
-- python test.py [options] 
-- python gradio_demo.py [interactive options]

--img_dir                Input folder.
--save_dir               Output folder.
--upscale                Upsampling ratio of given inputs. Default: 1
--SUPIR_sign             Model selection. Default: 'Q'; Options: ['F', 'Q']
--seed                   Random seed. Default: 1234
--min_size               Minimum resolution of output images. Default: 1024
--edm_steps              Numb of steps for EDM Sampling Scheduler. Default: 50
--s_stage1               Control Strength of Stage1. Default: -1 (negative means invalid)
--s_churn                Original hy-param of EDM. Default: 5
--s_noise                Original hy-param of EDM. Default: 1.003
--s_cfg                  Classifier-free guidance scale for prompts. Default: 7.5
--s_stage2               Control Strength of Stage2. Default: 1.0
--num_samples            Number of samples for each input. Default: 1
--a_prompt               Additive positive prompt for all inputs. 
    Default: 'Cinematic, High Contrast, highly detailed, taken using a Canon EOS R camera, 
    hyper detailed photo - realistic maximum detail, 32k, Color Grading, ultra HD, extreme
     meticulous detailing, skin pore detailing, hyper sharpness, perfect without deformations.'
--n_prompt               Fixed negative prompt for all inputs. 
    Default: 'painting, oil painting, illustration, drawing, art, sketch, oil painting, 
    cartoon, CG Style, 3D render, unreal engine, blurring, dirty, messy, worst quality, 
    low quality, frames, watermark, signature, jpeg artifacts, deformed, lowres, over-smooth'
--color_fix_type         Color Fixing Type. Default: 'Wavelet'; Options: ['None', 'AdaIn', 'Wavelet']
--linear_CFG             Linearly (with sigma) increase CFG from 'spt_linear_CFG' to s_cfg. Default: False
--linear_s_stage2        Linearly (with sigma) increase s_stage2 from 'spt_linear_s_stage2' to s_stage2. Default: False
--spt_linear_CFG         Start point of linearly increasing CFG. Default: 1.0
--spt_linear_s_stage2    Start point of linearly increasing s_stage2. Default: 0.0
--ae_dtype               Inference data type of AutoEncoder. Default: 'bf16'; Options: ['fp32', 'bf16']
--diff_dtype             Inference data type of Diffusion. Default: 'fp16'; Options: ['fp32', 'fp16', 'bf16']
```

### Python Script
```Shell
# Seek for best quality for most cases
CUDA_VISIBLE_DEVICES=0,1 python test.py --img_dir '/opt/data/private/LV_Dataset/DiffGLV-Test-All/RealPhoto60/LQ' --save_dir ./results-Q --SUPIR_sign Q --upscale 2
# for light degradation and high fidelity
CUDA_VISIBLE_DEVICES=0,1 python test.py --img_dir '/opt/data/private/LV_Dataset/DiffGLV-Test-All/RealPhoto60/LQ' --save_dir ./results-F --SUPIR_sign F --upscale 2 --s_cfg 4.0 --linear_CFG
```

### Gradio Demo
```Shell
CUDA_VISIBLE_DEVICES=0,1 python gradio_demo.py --ip 0.0.0.0 --port 6688
```

### Online Demo (Coming Soon)



---

## BibTeX
    @misc{yu2024scaling,
      title={Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic Image Restoration In the Wild}, 
      author={Fanghua Yu and Jinjin Gu and Zheyuan Li and Jinfan Hu and Xiangtao Kong and Xintao Wang and Jingwen He and Yu Qiao and Chao Dong},
      year={2024},
      eprint={2401.13627},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
    }

## 📧 Contact
If you have any question, please email `fanghuayu96@gmail.com`.