Subtitle Generation Project

This project provides an end-to-end pipeline for generating subtitles from video files. It includes preprocessing, subtitle extraction, and a web interface for user interaction.

How to Run

1. Using Docker (Recommended)

1. Build the Docker image:

   docker build -t subtitle_generation .

2. Run the container:

   docker run -p 8000:8000 -p 8501:8501 subtitle_generation

   • Port 8000: FastAPI backend
   • Port 8501: Streamlit frontend

3. To check where files are saved in DOCKER ```bash docker exec -it <container_name> bash

   Where you can see all the files.
   
   

2. Running Locally (Without Docker)

1. Install Python 3.10+ and pip
2. Install dependencies:

   pip install torch torchvision torchaudio
   pip install -r requirement.txt

3. Start the application:

   bash start.sh

---

File and Folder Structure

Top-Level Files

• dockerfile: Docker configuration to build and run the project in a containerized environment.
• front_end.py: Likely contains the Streamlit web interface for user interaction.
• main.py: Main entry point for the FastAPI backend server.
• requirement.txt: Lists all Python dependencies required for the project.
• start.sh: Shell script to start both backend and frontend services.
• readme.md: This documentation file.

Folders

preprocess/

Contains core Python scripts for processing videos and generating subtitles:

• full_movie_sub.py: Handles subtitle generation for full-length movies.
• functions.py: Utility functions used across the project.
• mail.py: Handles email notifications or sending results.
• subtitle.py: Main logic for subtitle extraction from video/audio.
• update_srt.py: Updates or merges SRT subtitle files.

uploads/

Stores uploaded video files and generated subtitles, organized by unique session IDs:

• <session_id>/: Each session has its own folder.
  • chunk/: Contains video chunks (e.g., 1.mp4, 2.mp4, ...).
  • original_file/: Stores the original uploaded video file.
  • srt/: Contains generated subtitle files (e.g., 1.srt, 2.srt, full.srt).

---

Notes

• The project exposes two ports: 8000 (FastAPI backend) and 8501 (Streamlit frontend).
• Make sure to have all dependencies installed as specified in requirement.txt.
• For any issues, check the logs