manga-translator-ui  by hgmzhn

Open-source tool for automatic manga translation and re-captioning

Created 4 months ago
488 stars

Top 63.2% on SourcePulse

GitHubView on GitHub
Project Summary

This project offers an open-source tool for automatically translating text within manga images. It targets comic enthusiasts and creators, providing automated detection, translation via multiple AI engines, and a visual editor for fine-tuning, streamlining localization.

How It Works

Leveraging the manga-image-translator core, it integrates OCR (PaddleOCR) and AI translation services (OpenAI, Gemini, Sakura). Intelligent typesetting adjusts translated text within panels, complemented by a visual editor for manual adjustments, enabling flexible automated or semi-manual localization.

Quick Start & Requirements

  • Installation: A recommended one-click installation script automates environment setup (Miniconda, Python 3.12, Git) and dependency installation, including GPU detection (NVIDIA CUDA 12.x, experimental AMD ROCm for RX 7000/9000). Pre-built CPU/NVIDIA GPU packages and experimental Docker are alternatives. Source installation requires Python 3.12.
  • Prerequisites: NVIDIA GPU with CUDA 12.x (driver >= 525.60.13) for GPU acceleration; CPU fallback available. AMD GPU support is experimental and limited. API keys needed for OpenAI/Gemini.
  • Links: Online Docs, GitHub Releases.

Highlighted Details

  • Supports Japanese, Korean, and American comics (B&W and color).
  • Offers 5 translation engines, including high-quality multimodal AI (GPT-4o, Gemini).
  • Visual editor allows precise control over text boxes, styling, and masking.
  • Automated workflow: detection, OCR, translation, typesetting.

Maintenance & Community

Actively developed with a QQ group (1074238546) for support. Builds on manga-image-translator and integrates models from PaddleOCR, bilibili/ailab, and lhj5426/YSG.

Licensing & Compatibility

Released under GPL-3.0. This copyleft license may restrict integration into closed-source commercial products.

Limitations & Caveats

AMD GPU support via ROCm is experimental, limited to RX 7000/9000 series (script install only); packaged versions lack AMD support. Docker deployment is experimental. High-quality AI translation requires external API keys.

Health Check
Last Commit

1 day ago

Responsiveness

Inactive

Pull Requests (30d)
3
Issues (30d)
15
Star History
259 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.