Skip to content
VANHONG

Mobile · Web case study

DramTranslatorApp

Automated video subtitle extraction and translation pipeline.

Project overview

Multilingual drama translation pipeline with Faster Whisper and FFmpeg.

Automated video subtitle extraction and translation pipeline.

Problem or motivation

Extracting, transcribing, translating, and hardcoding foreign video subtitles into Khmer requires hours of tedious manual alignment.

What I built

  • Automated audio extraction and speech recognition pipeline orchestrated with Faster Whisper and OpenCV text recognition.
  • Asynchronous subtitle timing synchronization and video multiplexing using FFmpeg and background task queues.
  • FastAPI microservice backend paired with a cross-platform companion interface.

Technology stack

  • Dart
  • Python (FastAPI)
  • Faster Whisper
  • FFmpeg
  • Next.js
  • Docker
  • Full stack repository featuring Dart, Python, and TypeScript.

Screenshots

No screenshot has been added yet; the cover above is a typographic placeholder, not a real screenshot.

What I learned

  • Orchestrating containerized compute pipelines with Docker Compose for intensive video processing jobs.
  • Handling audio drift, subtitle timestamps, and variable-framerate media alignment.

There is no live deployment for this project.