中文

Comic Translate

Comic Translate is a project created to utilize Large Language Models like GPT to translate comics.

Verified

Core facts

Read by a script from the official API, not rewritten by a model.

License
Apache-2.0
Languages
Python
Maintainer
ogkalu2
Verified

What it is

Comic Translate is a project created to utilize Large Language Models (LLMs) like GPT to translate comics. It currently supports translating comics from English, Korean, Japanese, French, Simplified Chinese, Traditional Chinese, Russian, German, Dutch, Spanish and Italian. It can translate into those languages and more.

Main features

  1. Translation currently supports using GPT-4.1 and Claude-4.5.

  2. The LLMs used are fed the entire page text to aid translations.

  3. There is also an option to provide the image itself for further context.

  4. Version 2.0 introduces a Manual Mode, and when issues occur with Automatic Mode (such as no text detected, incorrect OCR or insufficient cleaning), corrections can now be made.

  5. In Automatic Mode, once an image has been processed, it is loaded in the Viewer or stored to be loaded on switch, so reading in the app can continue while the other images are being translated.

  6. An RT-DETR-v2 model is trained on 11k images of comics (Manga, Webtoons, Western).

Getting started

  1. The desktop app can be downloaded and installed for Windows and macOS.

  2. A Comic Translate extension can be installed for Chromium-based browsers (Chrome, Edge, Brave, etc.).

  3. To run the source code directly, install Python 3.12.

  4. Running from source also involves installing git and uv.

  5. Running a command in the comic-translate directory launches the GUI.

  6. To update, run a command in the comic-translate folder.

From the README · bash
https://www.python.org/downloads/
From the README · bash
https://git-scm.com/
From the README
https://docs.astral.sh/uv/getting-started/installation/
From the README · bash
git clone https://github.com/ogkalu2/comic-translate
cd comic-translate
uv init --python 3.12
From the README · bash
uv add -r requirements.txt --compile-bytecode
From the README · bash
uv run comic.py

Who it is for

GPU acceleration is currently only available when running from source. With an NVIDIA GPU, the project recommends running a command. During Python setup, tick Add python.exe to PATH.

On Windows, Smart Screen is ignored by clicking More info and then Run anyway. On macOS, after trying to open the app, click Open Anyway under Privacy and Security in Settings. CBR/RAR extraction requires a supported helper such as UnRAR, Unar, or 7-Zip.

The selected font must support characters of the target language. Ctrl plus the mouse wheel zooms, and otherwise the wheel scrolls vertically. The Right and Left keys navigate between images.

The usual trackpad gestures work for viewing the image. Segmentation is algorithmic and based on the boxes provided by the detection model. Inpainting removes the segmented text.

Text rendering wraps text in bounding boxes obtained from bubbles and text.

Sources

This page is written from the official sources below. Open them to check the original text.

Verified ·

This page is written by AI from the official sources above and checked automatically. This site has not tested it. Found a mistake? Tell us.