update

README.en.md

@@ -14,13 +14,13 @@
 
 </div>
 
-## 🌟 What Can It Do
+## 🌟 Our Vision
 
-- 🍖 Fully automated video translation, generating Netflix-quality subtitles!
+VideoLingo is an all-in-one video translation and localization tool designed to generate Netflix-quality subtitles. It aims to eliminate stiff machine translations and multi-line subtitles, enabling knowledge sharing across language barriers worldwide.
 
-- 🎤 Clone your own voice for dubbing! (🚧 Still in development)
+VideoLingo automates the entire process of video downloading, audio extraction, high-precision speech recognition, subtitle generation, high-quality text translation, subtitle alignment, and seamless integration of translated subtitles into the original video. With an intuitive Streamlit web interface, you can complete the entire workflow with just a few clicks, easily creating localized videos with Netflix-quality subtitles.
 
-- ✨ Click-and-done in Streamlit!
+VideoLingo is also actively developing voice cloning technology, which will soon support video dubbing, further enhancing the localization experience. Whether you're a content creator, educator, or someone with multilingual communication needs, VideoLingo can be a powerful assistant, helping you overcome language barriers and connect with global audiences.
 
 > Check out the results! 💪
 
@@ -41,20 +41,22 @@ https://github.com/user-attachments/assets/25264b5b-6931-4d39-948c-5a1e4ce42fa7
 
 ## ✨ Features
 
-- Subtitle segmentation using NLP and LLM
+- Word-level timeline subtitle recognition using WhisperX
 
-- Intelligent terminology knowledge base for context-aware translation
+- Subtitle segmentation based on sentence meaning using NLP and GPT
 
-- Three-step translation process: Direct Translation - Reflection - Improvement
+- GPT-summarized intelligent terminology knowledge base for context-aware translation
 
-- Precise word-level subtitle alignment
+- Three-step translation process: Direct Translation, Reflection, and Paraphrasing to eliminate awkward machine translations
 
-- Create 5 minutes of Netflix-quality bilingual subtitles for just 1 yuan
+- Netflix-level single-line subtitle length and translation quality
+
+- One-click integrated package launch, one-click video production in Streamlit!
 
 - Developer-friendly: Step-by-step structured files for easy customization: [Chinese Technical Documentation](./docs/README_guide_zh.md) | [English Technical Documentation](./docs/README_guide_en.md) 
     > You can even run each `step__.py` file under `core` individually!   
 
-## 🏠 [Local Deployment Guide](./docs/install_locally_en.md)
+## 🏠 [Local Installation Guide](./docs/install_locally_en.md)
 
 ## 🚧 Current Limitations
 
@@ -83,13 +85,9 @@ We are continuously improving VideoLingo, but there are still some limitations:
 
 Thanks to the following open-source projects for their contributions:
 
-- [whisper](https://github.com/openai/whisper): OpenAI's open-source automatic speech recognition system
 - [whisper-timestamped](https://github.com/linto-ai/whisper-timestamped): Extension adding timestamp functionality to Whisper
 - [whisperX](https://github.com/m-bain/whisperX): Extension adding timestamp functionality to Whisper
 - [yt-dlp](https://github.com/yt-dlp/yt-dlp): Command-line tool for downloading YouTube videos and content from other websites
-- [GPT-SoVITS](https://github.com/RVC-Project/GPT-SoVITS) & [GPT-SoVITS-Inference](https://github.com/X-T-E-R/GPT-SoVITS-Inference): Speech synthesis system and inference library based on GPT and SoVITS
-- [FFmpeg](https://github.com/FFmpeg/FFmpeg): Complete, cross-platform solution for handling multimedia content
-- [Ultimate Vocal Remover GUI v5 (UVR5)](https://github.com/Anjok07/ultimatevocalremovergui): Tool for separating vocals and instrumentals in music
 - [json_repair](https://github.com/mangiucugna/json_repair): Super powerful library for repairing and parsing GPT's JSON output, seamlessly replacing json.loads
 
 ## 🤝 Contributions Welcome

README.md

@@ -14,13 +14,13 @@
 
 </div>
 
-## 🌟 能做什么
+## 🌟 初衷
 
-- 🍖 全自动视频搬运工，生成 Netflix 品质的字幕！
+VideoLingo 是一站式视频翻译本地化工具，旨在生成 Netflix 级别的高质量字幕，告别生硬机翻，告别多行字幕，让全世界的知识能够跨越语言的障碍共享。
 
-- 🎤 克隆自己的声音进行配音!（🚧 仍在开发中）
+VideoLingo 能够自动下载视频、提取音频、进行高精度的语音识别、生成字幕、执行高质量的文本翻译和字幕对齐，并将翻译后的字幕无缝集成到原始视频中。通过直观的 Streamlit 网页界面，只需点击几下就能完成整个流程，轻松创建出具有 Netflix 品质字幕的本地化视频。
 
-- ✨ 在 streamlit 中点击-完成！
+VideoLingo 还在积极开发声音克隆技术，很快将支持视频配音，进一步提升视频的本地化体验。无论是内容创作者、教育工作者还是多语言传播需求者，VideoLingo 都能成为强大的助手，帮助跨越语言障碍，连接全球观众。
 
 > 看看效果吧！💪
 
@@ -41,21 +41,23 @@ https://github.com/user-attachments/assets/25264b5b-6931-4d39-948c-5a1e4ce42fa7
 
 ## ✨ 特点
 
-- 使用 NLP 和 LLM 进行字幕分割
+- 使用 WhisperX 进行单词级时间轴字幕识别
 
-- 智能术语知识库，实现上下文感知翻译
+- 使用 NLP 和 GPT 根据句意进行字幕分割
 
-- 三步翻译过程：直接翻译 - 反思 - 改进
+- GPT 总结智能术语知识库，实现上下文感知翻译
 
-- 精确的单词级字幕对齐
+- 三步直译、反思、意译，告别诡异机翻
 
-- 仅需 1 元即可创作 5 分钟的 Netflix 级双语字幕
+- Netflix 级别的单行字幕长度与翻译质量
+
+- 一键整合包启动，在 streamlit 中一键出片！
 
 - 开发者友好：逐步结构化文件，便于自定义开发 : [中文技术文档](./docs/README_guide_zh.md) | [英文技术文档](./docs/README_guide_en.md) 
     > 你甚至可以单独运行每一个`core`下的`step__.py`文件！   
 
 
-## 🏠 [本地部署教程](./docs/install_locally_zh.md)
+## 🏠 [本地安装教程](./docs/install_locally_zh.md)
 
 ## 🚧 当前限制
 
@@ -82,15 +84,11 @@ https://github.com/user-attachments/assets/25264b5b-6931-4d39-948c-5a1e4ce42fa7
 
 ## 🙏 致谢
 
-感谢以下开源项目的贡献:
+感谢以下开源项目的贡献：
 
-- [whisper](https://github.com/openai/whisper): OpenAI的开源自动语音识别系统
 - [whisper-timestamped](https://github.com/linto-ai/whisper-timestamped): 为Whisper添加时间戳功能的扩展
 - [whisperX](https://github.com/m-bain/whisperX): 为Whisper添加时间戳功能的扩展
 - [yt-dlp](https://github.com/yt-dlp/yt-dlp): 用于下载YouTube视频和其他网站内容的命令行工具
-- [GPT-SoVITS](https://github.com/RVC-Project/GPT-SoVITS) & [GPT-SoVITS-Inference](https://github.com/X-T-E-R/GPT-SoVITS-Inference): 基于GPT和SoVITS的语音合成系统及推理库
-- [FFmpeg](https://github.com/FFmpeg/FFmpeg): 用于处理多媒体内容的完整多平台解决方案
-- [Ultimate Vocal Remover GUI v5 (UVR5)](https://github.com/Anjok07/ultimatevocalremovergui): 用于分离音乐中的人声和伴奏的工具
 - [json_repair](https://github.com/mangiucugna/json_repair): 超无敌的 修复解析 gpt 的 json 输出的库，无缝替代 json.loads
 
 ## 🤝 欢迎贡献