Multilingual Video Dubbing (Prototype)

FastAPI demo: English video → Hindi/Marathi dubbed outputs via Whisper, Google Translate, gTTS.

Prerequisites

• ffmpeg on PATH
• Python 3.10+

Install

python -m venv .venv
. .venv/Scripts/activate
pip install -r requirements.txt

Run

uvicorn app:app --reload --host 0.0.0.0 --port 8000

Languages

GET /languages returns supported language codes and names. Defaults for dubbing include: hi, mr, bn, ta, te. Also available: gu, kn, ml, pa.

API

• POST /upload-video: form-data file
• POST /dub-video: form-data video_id, target_languages (can include comma-separated codes, e.g. hi,mr,bn,ta,te), model_size, generate_subtitles

Outputs are served under /outputs/... for preview.

Notes: Whisper downloads on first use; googletrans/gTTS need internet; timing fit is naive but OK for demo.

Docker

Build and run the API in a container with ffmpeg included:

docker build -t multilingual-dubbing:latest .
docker run --rm -p 8000:8000 -v %cd%/data:/app/data multilingual-dubbing:latest

Linux/macOS volume flag:

docker run --rm -p 8000:8000 -v $(pwd)/data:/app/data multilingual-dubbing:latest

Visit http://localhost:8000/ui and http://localhost:8000/outputs.