GIMP3-ML

Machine Learning plugins for GIMP 3.

Forked from forked version of the original version to improve the user experience in several aspects:

(mine) adapted to run on Apple M1 and with MPS backend (check this)
Added more models.
Models are run with Python 3.10+.
Full error text is shown in the GIMP error dialog and in debug console.
Additional alpha channel handling in some plugins.
Automatic installation for Windows systems.
And other smaller improvements.

The plugins have been tested with GIMP 2.99.14 on the following systems:

Linux Ubuntu 22

Installation Steps

Install GIMP3 for macOS.
Download/clone this repository.
Edit ...
You should now find the GIMP-ML plugins under Layers → GIMP-ML.
You can download the weights here, or from the weight links below.

Original text from the forked repository:

References

Background Removal

Source: https://github.com/danielgatis/rembg
Weights:
- u2net (download, source): A pre-trained model for general use cases.
- u2netp (download, source): A lightweight version of u2net model.
- u2net_human_seg (download, source): A pre-trained model for human segmentation.
- (unused) u2net_cloth_seg (download, source): A pre-trained model for Cloths Parsing from human portrait. Here clothes are parsed into 3 category: Upper body, Lower body and Full body.
License: MIT License

Anime-style Inpainting

Source: https://github.com/youyuge34/Anime-InPainting
Weights: Google Drive | Baidu
License: Creative Commons Attribution-NonCommercial 4.0 International

@inproceedings{nazeri2019edgeconnect,
  title={EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning},
  author={Nazeri, Kamyar and Ng, Eric and Joseph, Tony and Qureshi, Faisal and Ebrahimi, Mehran},
  journal={arXiv preprint},
  year={2019}}

Demosaics

Source:
- Demosaics: https://github.com/rekaXua/demosaic_project
- ESRGAN: https://github.com/xinntao/ESRGAN
Weights: 4x_FatalPixels
Licenses:
- Demosaics: GNU Affero General Public License v3.0
- ESRGAN: Apache-2.0 license

[Paper]
Xintao Wang, Liangbin Xie, Chao Dong, Ying Shan
Applied Research Center (ARC), Tencent PCG
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences

Inpainting

@article{Moskalenko_2020,
	doi = {10.51130/graphicon-2020-2-4-18},
	url = {https://doi.org/10.51130%2Fgraphicon-2020-2-4-18},
	year = 2020,
	month = {dec},
	pages = {short18--1--short18--9},
	author = {Andrey Moskalenko and Mikhail Erofeev and Dmitriy Vatolin},
	title = {Deep Two-Stage High-Resolution Image Inpainting},
	journal = {Proceedings of the 30th International Conference on Computer Graphics and Machine Vision ({GraphiCon} 2020). Part 2}}

SRResNet

Source: https://github.com/twtygqyy/pytorch-SRResNet
Torch Hub fork: https://github.com/valgur/pytorch-SRResNet
License: MIT
C. Ledig et al., “Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 105–114.

Neural Colorization

Source: https://github.com/zeruniverse/neural-colorization
Torch Hub fork: https://github.com/valgur/neural-colorization
License:
- GNU GPL 3.0 for personal or research use
- Commercial use prohibited
- Model weights released under CC BY 4.0
Based on fast-neural-style:
- https://github.com/jcjohnson/fast-neural-style
- License:
  - Free for personal or research use
  - For commercial use please contact the authors
- J. Johnson, A. Alahi, and L. Fei-Fei, “Perceptual Losses for Real-Time Style Transfer and Super-Resolution,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9906 LNCS, 2016, pp. 694–711.

Edge Detection (DexiNed)

Source: https://github.com/xavysp/DexiNed
Weights: BIPED
License: MIT license

@misc{soria2021dexined_ext,
    title={Dense Extreme Inception Network for Edge Detection},
    author={Xavier Soria and Angel Sappa and Patricio Humanante and Arash Arbarinia},
    year={2021},
    eprint={arXiv:2112.02250},
    archivePrefix={arXiv},
    primaryClass={cs.CV}}

DeblurGANv2

Source: https://github.com/TAMU-VITA/DeblurGANv2
Torch Hub fork: https://github.com/valgur/DeblurGANv2
License: BSD 3-clause
O. Kupyn, T. Martyniuk, J. Wu, and Z. Wang, “DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better,” in 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 2019, pp. 8877–8886.

Monodepth2

Source: https://github.com/nianticlabs/monodepth2
Torch Hub fork: https://github.com/valgur/monodepth2
C. Godard, O. Mac Aodha, M. Firman, and G. Brostow, “Digging Into Self-Supervised Monocular Depth Estimation,” in 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 2019, pp. 3827–3837.

Authors

UserUnknownFactor
Kritik Soman (kritiksoman) – original GIMP-ML implementation

License

MIT

Please note that additional license terms apply for each individual model. See the references list for details. Many of the models restrict usage to non-commercial or research purposes only.

ivvv/GIMP3-ML