LiveVTT

Subtitle generation tools for live and archived streams. LiveVTT focuses on producing WebVTT/TTML assets and SMIL manifests for downstream players and packaging workflows.

🚀 Quick Start

Requirements

• Python 3.10+
• uv CLI: used to sync and install dependencies.
  • Install via pip install uv for Python environments.
  • Or use brew install uv on macOS / Linux if available.
  • See the official uv project page for more installation options: https://uv.run
• FFmpeg + FFprobe: FFprobe is included with FFmpeg. Download from the official FFmpeg site. Install via apt install ffmpeg (Linux), brew install ffmpeg (macOS), or choco install ffmpeg (Windows).

Install dependencies

uv sync

Transcribe archived media

python src/python/tools/archive_transcriber.py /path/to/media --max-files 1 --progress  # Example: transcribe media in directory, process up to 1 file with progress display

Run the polling service

python src/python/services/subtitle_autogen.py /path/to/watch --batch-size 5 --interval 300  # Example: watch directory for new files, batch size 5, poll every 300 seconds

🛠️ Core Tools

archive_transcriber

Batch transcription of archived broadcast chunks with bilingual WebVTT output and SMIL manifest generation.

subtitle_autogen

Polling service for automated transcription + SMIL regeneration.

test_integration

System health check for the local toolchain.

📁 Project Structure

livevtt/
├── docs/                          # Documentation
│   └── DEMO.md                    # Demo instructions
├── src/                           # Source code
│   └── python/                    # Python utilities
│       ├── services/
│       ├── tools/
│       └── utils/
├── config/                        # Configuration files
│   ├── vocabulary.json
│   └── filter.json
├── deploy/                        # Deployment files
├── test_integration
└── main.py                        # Live transcription pipeline