Translumo is an advanced real-time screen translator developed by Danily07. The service is capable of detecting and translating text that appears in a selected area in real-time, for example, subtitles in videos or games, hardcoded subtitles, static text, and etc.
The primary features of Translumo include:
- High text recognition precision: The service can combine several OCR engines simultaneously, allowing it to accurately identify text. It employs a machine learning training model to score each recognized result by OCR and selects the best one.
- Simple Interface: Translumo was designed with a primary objective to create a tool that doesn’t require manual adjustments for each case and is convenient for everyday use.
- Low Latency: Translumo has implemented several optimizations to minimize system performance impact and reduce the latency between the moment a text appears and its actual translation.
Translumo integrates modern OCR engines including Tesseract 5.2, WindowsOCR, and EasyOCR. It supports multiple translators such as Google Translate, Yandex translate, Naver Papago, and DeepL. The service can recognize several languages including English, Russian, Japanese, Chinese (simplified), and Korean. It can translate into languages such as English, Russian, Japanese, Chinese (simplified), Korean, French, Spanish, German, Portuguese, and Italian.
The system requirements for running Translumo include Windows 10 build 19041 (20H1) / Windows 11, DirectX11, and 8GB RAM.
The service is compatible with the OpenCvSharp4 framework, a wrapper for .NET for OpenCV. This indicates that Translumo likely uses OpenCV for some of its image processing and computer vision tasks.