README-EN.md

AI Auto Video(Audio) Translation

Chenyme-AAVT V0.8.3

Thank you very much for coming to my Automatic Video Translation project! This project aims to provide a simple and easy-to-use automatic video (audio) recognition and translation tool to help you quickly recognize video subtitles and generate subtitle files, and then merge the translated subtitles with the original video for quick video translation.

• Note0: The subtitle translation misalignment issue will be gradually optimized. Due to postgraduate studies, the update speed may slow down, thank you for your understanding~~~
• Note1: It is recommended to use the Faster-whisper and Large models for the best sentence breaking and recognition experience!
• Note2: The new version has significant changes and many bugs, so updates are frequent recently, it is recommended to update!
• Note3: After this version stabilizes, updates will slow down, studies are important, if you have any questions, you can join the group to discuss!

This update really took a long time! Give a free star to encourage it~ Thank you! AAVT Project Documentation

Test effect click to download

Project Highlights

• Supports OpenAI API interface calls and Faster-Whisper local operation.
• Supports GPU acceleration, VAD assistance.
• Supports various translation modes such as ChatGPT, KIMI, DeepSeek, ChatGLM, locally deployed models.
• Supports adjusting various parameters to meet customized needs.
• Supports recognition and translation of multiple languages and multiple file formats.
• Supports one-click generation of processed content.
• Supports subtitle modification, fine-tuning, preview.
• Supports direct AI summary, Q&A of content.
• Supports direct video generation of graphic blog posts.

How to Install

1. Install Python

• Please ensure that the Python version is greater than 3.8

2. Install FFmpeg

• The Full version in the Release already includes the FFmpeg library
• Set the FFmpeg environment variable
  • Use the Win+R shortcut to open the Run dialog box.
  • Enter rundll32 sysdm.cpl,EditEnvironmentVariables.
  • In User variables, find Path.
  • Click New and enter the path to FFmpeg. Example: D:\APP\ffmpeg\bin (please adjust according to your actual path).

3. Run install.bat

• Choose the corresponding version of install.bat and wait for all dependencies to be installed.
• If running on CPU, select the CPU version, similarly for CUDA11.8, CUDA12.1.

TODO

Recognition Related

• Replace with a faster Whisper project
• Support local model loading
• Support personal fine-tuning of Whisper models
• VAD assistance optimization
• Word-level sentence breaking optimization
• More language recognition

Translation Related

• Translation optimization
• More language translations
• More translation models
• More translation engines
• Support local large language model translation

Subtitle Related

• Personalized subtitles
• More subtitle formats
• Subtitle preview, real-time modification
• Automated subtitle text proofreading
• Dual subtitles

Other

• Video summary, listing key points
• Video preview
• AI assistant
• Video generation of blog posts*
• Real-time voice translation
• Video Chinese dubbing

Note: Features marked with * are still unstable and may have some bugs.

Star History

Project Interface Preview

Main Page

Settings

Video Recognition

Parameter Settings

Running Interface

Video Generation

Subtitle Fine-tuning

Content Assistant

Parameter Settings

Running Interface

Subtitle Translation

Video Blog

![09f60b8099f8ce19b83f4da63b26817](https://github.com/Chenyme/Chenyme-AAVT/assets/118253778/bbfca353-53d4-4a19-994f-7beddbbf17