teamstudy18-maker/SomaliScribeAI

AI-powered Somali subtitle generator # SomaliScribe AI An AI-powered web app that generates Somali subtitles from audio and video files. Upload media, edit subtitles in the browser, and export SRT, VTT, or a video with burned-in captions. ## Features - **Web interface**: Dark-themed UI (SomaliScribe AI) with Home, Upload, Editor, and History - **User management**: Register, login, guest mode, role-based access (User, Admin) - **Media upload**: Audio (MP3, WAV, M4A, FLAC, OGG, AAC) and video (MP4, MKV, AVI). File size limit 100MB. Upload progress indicator - **Somali ASR**: Automatic speech recognition using Google Speech Recognition (so-SO) with dataset fallback (6,751 Somali audio entries) - **Subtitle generation**: Time-aligned segments; supports SRT and VTT - **Subtitle editor**: Edit text, adjust timestamps, merge/split segments, live caption preview with customizable font and colors - **Export & download**: Download SRT or VTT; burn subtitles into video (or generate a video from audio + subtitles) - **History**: Per-user job history with links to edit, export, and burn-in - **Admin panel**: Manage users, files, and view transcription statistics ## Dataset Structure ``` dataset/ ├── metadata.csv # Audio entries with transcriptions ├── train.csv ├── validation.csv ├── test.csv └── wavs/ # Audio files (.wav) ``` ## Installation 1. **Python**: 3.7 or higher 2. **Create and activate a virtual environment** (recommended): ```bash python -m venv .venv .venv\Scripts\activate # Windows # source .venv/bin/activate # Linux/macOS ``` 3. **Install dependencies**: ```bash pip install -r requirements.txt ``` 4. **PostgreSQL**: Install PostgreSQL (v16+) and create the database: ```bash psql -U postgres CREATE DATABASE somali_subtitles; \q ``` By default the app connects to `postgresql://postgres:postgres@localhost:5432/somali_subtitles`. Override with the `DATABASE_URI` environment variable: ```bash set DATABASE_URI=postgresql://user:pass@host:5432/dbname ``` 5. **ffmpeg** (for …