The project's goal was to develop a system that could automatically dub any English video into Hindi or any other Indian regional language while ensuring the output works both online and offline. The challenge was to build a model that could seamlessly synchronize audio and lip movements while maintaining a user-friendly interface. Additionally, the solution had to be optimized for performance, producing fast results and supporting offline usage to make it accessible in low-connectivity environments.

Our solution addresses the problem by utilizing machine learning models and audio processing techniques to convert English audio from videos into selected Indian languages. The system performs real-time synchronization of dubbed audio with the original video, maintaining lip-sync accuracy. It can process YouTube video URLs and convert them into dubbed versions efficiently. The project includes an easy-to-use interface developed in Python with Gradio, allowing users to generate dubbed videos quickly, even in offline mode.

• Offline Support: Works without an internet connection, ensuring accessibility in low-connectivity areas.
• YouTube Video Dubbing: Converts English YouTube video URLs into dubbed videos in Hindi or other Indian languages.
• Audio Synchronization: Ensures that the dubbed audio is synchronized perfectly with the original video.
• Lip Sync: Maintains lip synchronization for a realistic dubbing experience.
• Fast Output: Provides quick results through an optimized Gradio Python UI.

• Python: For the core application logic and machine learning model integration.
• Machine Learning (ML): To handle speech recognition, translation, and text-to-speech synthesis.
• REST API: For handling requests and integrating external services.
• Audio Processing Libraries: For audio synchronization and lip sync.
• Gradio: For building a simple and interactive user interface.