I'm the admin of a wiki where my team translates webcomics in a very specific way. Mostly on the subject of furry, there are sections about Simba's world, wolves, little ponies, also there is science fiction and comics about computer games. Our license is CC BY-NC-SA, the freest available in our field. Despite our fan uncommercial status, we try to reach out to every author whose work we take into translation, and we put in the descriptions all the author's financial references we can find.
Original, permission to distribute is granted by the author
In order to preserve our collective nature, we do not use Photoshop but a specially adapted web-editor called CoTAN (comic translator by applicable notes). The editor operates with two types of stickers - background color and text on a transparent background. All the rest - rotation, size, color, font - is available. And the base font is not Comic Sans! Theoretically, the author can even post a retouched drawing on their own and put text where they like.
Our version, is also used as the basis for translation into other languages Also, it is available for reading by special programs for people with physical disabilities
This markup architecture has allowed for extensive "clonable" multilingualism, but I have so far settled on 2 dozen languages, including such artificial ones as Sindarin and Esperanto. Our main auto-translator is the convolutional neural network DeepL, which also helped to write this text. Additional translators are Google Translate and Yandex Translate. Nevertheless, I am making an effort to involve native speakers. The main goal of my project I see as removing cultural boundaries in the field of webcomics, seriously.
The result can be output in a finished form, photographed by a special bot. Also, if the comic was published quite a long time ago and its current size is unsatisfactory, and the sources are lost, you can pass it through the Waifu2x-Caffe neural network, which enlarges and cleans it of compression artifacts. After that we not only allow the authors to reuse the result, but even recommend doing so.
Eventually I envision us evolving into a compact group of high-performance neural networks for recognizing, translating, publishing, and distributing comics, but for now my project can be considered a technical assignment for that.