Spaces:
Running
on
Zero
Running
on
Zero
File size: 2,782 Bytes
291fb27 e9d7349 3e8a194 291fb27 e9d7349 291fb27 c509e76 e9d7349 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 |
---
title: Documents Restoration
emoji: 📚
colorFrom: purple
colorTo: indigo
sdk: gradio
sdk_version: 4.31.0
app_file: app.py
pinned: false
short_description: Enhance photo of a document with selected approaches!
---
<div align=center>
# DocRes: A Generalist Model Toward Unifying Document Image Restoration Tasks
</div>
<p align="center">
<img src="images/motivation.jpg" width="400">
</p>
This is the official implementation of our paper [DocRes: A Generalist Model Toward Unifying Document Image Restoration Tasks](https://arxiv.org/abs/2405.04408).
## News
🔥 A comprehensive [Recommendation for Document Image Processing](https://github.com/ZZZHANG-jx/Recommendations-Document-Image-Processing) is available.
## Inference
1. Put MBD model weights [mbd.pkl](https://1drv.ms/f/s!Ak15mSdV3Wy4iahoKckhDPVP5e2Czw?e=iClwdK) to `./data/MBD/checkpoint/`
2. Put DocRes model weights [docres.pkl](https://1drv.ms/f/s!Ak15mSdV3Wy4iahoKckhDPVP5e2Czw?e=iClwdK) to `./checkpoints/`
3. Run the following script and the results will be saved in `./restorted/`. We have provided some distorted examples in `./input/`.
```bash
python inference.py --im_path ./input/for_dewarping.png --task dewarping --save_dtsprompt 1
```
- `--im_path`: the path of input document image
- `--task`: task that need to be executed, it must be one of _dewarping_, _deshadowing_, _appearance_, _deblurring_, _binarization_, or _end2end_
- `--save_dtsprompt`: whether to save the DTSPrompt
## Evaluation
1. Dataset preparation, see [dataset instruction](./data/README.md)
2. Put MBD model weights [mbd.pkl](https://1drv.ms/f/s!Ak15mSdV3Wy4iahoKckhDPVP5e2Czw?e=iClwdK) to `data/MBD/checkpoint/`
3. Put DocRes model weights [docres.pkl](https://1drv.ms/f/s!Ak15mSdV3Wy4iahoKckhDPVP5e2Czw?e=iClwdK) to `./checkpoints/`
2. Run the following script
```bash
python eval.py --dataset realdae
```
- `--dataset`: dataset that need to be evaluated, it can be set as _dir300_, _kligler_, _jung_, _osr_, _docunet\_docaligner_, _realdae_, _tdd_, and _dibco18_.
## Training
1. Dataset preparation, see [dataset instruction](./data/README.md)
2. Specify the datasets_setting within `train.py` based on your dataset path and experimental setting.
3. Run the following script
```bash
bash start_train.sh
```
## Citation:
```
@inproceedings{zhangdocres2024,
Author = {Jiaxin Zhang, Dezhi Peng, Chongyu Liu , Peirong Zhang and Lianwen Jin},
Booktitle = {In Proceedings of the IEEE/CV Conference on Computer Vision and Pattern Recognition},
Title = {DocRes: A Generalist Model Toward Unifying Document Image Restoration Tasks},
Year = {2024}}
```
## ⭐ Star Rising
[](https://star-history.com/#ZZZHANG-jx/DocRes&Timeline) |