Skip to content

Audio transcription with interactive HTML viewer featuring real-time word highlighting and translations

Notifications You must be signed in to change notification settings

shjala/syncscribe

Repository files navigation

SyncScribe 🎙️

SyncScribe takes an audio file and generates a standalone HTML viewer with embedded audio and synchronized transcription. Words are highlighted as they're spoken, and hovering over any word provides instant English translation. The generated file is fully self-contained and works completely offline, requiring no internet connection.

It uses local OpenAI's Whisper models via faster-whisper, if GPU-acceleration is available, it will be used. Checkout sample page for a demonstration.

Quick Start

Generate Interactive Transcription Viewer

# Basic usage
./create_audio_viewer.sh your_audio.wav

# With custom model
./create_audio_viewer.sh --model medium

Output: Standalone HTML file in out/ folder (or specified directory)

Requirements

  • Python 3.8+
  • NVIDIA GPU (recommended) with CUDA support

Installation

No manual installation needed, the scripts automatically:

  1. Create Python virtual environment
  2. Install all dependencies (faster-whisper, torch, librosa, etc.)
  3. Configure GPU/CUDA settings
  4. Download Whisper models to local cache
# Just clone and run
git clone <your-repo>
cd syncscribe
./create_audio_viewer.sh your_audio.wav

Everything is created and downloaded in the current directory for easy cleanup.

About

Audio transcription with interactive HTML viewer featuring real-time word highlighting and translations

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages