Arkwaifu is a website which, arranges and provides almost all artworks extracted from Arknights (the game). Arkwaifu also enlarges (4x) the artworks with super-resolution neural networks - Real-ESRGAN.
🎉 Arkwaifu v1 has released! Check it at arkwaifu.cc!
🇨🇳 The China mirror has been suspended due to the v0 → v1 upgrade. We will make our best efforts to get it back soon.
For more information of the frontend of Arkwaifu, please refer to the frontend repo.
- The arts are automatically updated as soon as the game updates.
- Assets are enlarged with super-resolution neural networks (Real-ESRGAN).
Now, only arts that appear in the "in-game stories", are available, including images, backgrounds and * characters*.
Images are the exquisite artworks that appear when some special events in the stories happen.
Backgrounds are the artworks that always appear during dialogue between characters.
Characters are the artworks that represents characters that appear during dialogue.
- Switchable language.
- Support searching.
Thanks to my friend Galvin Gao! He helped me a lot in the front-end development and choosing frameworks. I really appreciate the "getting hands dirty" methodology very much.
Thanks to my friend Martin Wang! He helped me in extracting the gamedata assets, and also in some details of automatically updating the assets from the game.
Thanks to my friend Lily! She drew the fascinating Phantom logo of this project.
Thanks to Penguin Statistics! The prototype of this project referenced and is inspired by Penguin Statistics' backend v3.
Thanks to xinntao, nihui, and the other contributors of Real-ESRGAN and Real-CUGAN! They created the neural networks this project utilizes for enlarging assets.
The source code of this project is licensed under the MIT License.
The assets of this project are licensed under Attribution-NonCommercial 4.0 International (CC BY-NC 4.0).
This project utilizes resources and other works from the game Arknights. The copyright of such works belongs to the provider of the game, 上海鹰角网络科技有限公司 (Shanghai Hypergryph Network Technology Co., Ltd).
This project utilizes Real-ESRGAN and Real-ESRGAN-ncnn-vulkan, which are respectively licensed under the BSD-3-Clause license and the MIT License.
This project utilizes Real-CUGAN and Real-CUGAN-ncnn-vulkan, which are both licensed under the MIT License.
Some initial template source code of this project is inspired by and copied from the backend v3 of Penguin Statistics, which is licensed under the MIT License.