kpttiling-wq/git-clone-https-github.com-lucy-kevin-Luganda-Speech-To-Text-API.git-cd-Luganda-Speech-To-Text-API

# Luganda Speech-to-Text API with Audio Cleaning Enhanced Luganda Speech-to-Text API featuring advanced audio cleaning, noise reduction, and voice enhancement for improved transcription accuracy. ## Features ### 🎙️ Core Features - **Real-time Luganda Speech-to-Text Transcription** using Wav2Vec2 model - **Professional Audio Cleaning Pipeline**: - 🔇 **Noise Reduction** - Removes background noise while preserving speech - 🔊 **Voice Enhancement** - High-pass filtering and dynamic compression - 📊 **Audio Normalization** - Consistent loudness levels - ✂️ **Silence Trimming** - Removes silence from beginning/end - **Multiple API Endpoints** for different use cases - **Comprehensive Audio Analysis** - Quality metrics and statistics - **Customizable Cleaning Parameters** - Fine-tune audio processing ## What's New in This Version ✨ **Audio Preprocessing**: Automatic noise reduction and voice enhancement before transcription ✨ **Quality Metrics**: Detailed audio statistics (RMS, peak amplitude, SNR, frequency analysis) ✨ **Flexible Configuration**: Control noise reduction strength and cleaning options ✨ **Multiple Endpoints**: Transcription, audio analysis, translation preparation ✨ **Production Ready**: Error handling, timeout management, configurable parameters ## Prerequisites - Python 3.7+ - pip package manager - 2GB+ RAM for model loading - Audio files in WAV, MP3, or FLAC format ## Installation ### 1. Clone the repository ```bash git clone github.com cd git-clone-https-github.com-lucy-kevin-Luganda-Speech-To-Text-API.git-cd-Luganda-Speech-To-Text-API ``` ### 2. Install dependencies ```bash pip install -r requirements.txt ``` ### 3. First Run The first time you run the API, it will download the pre-trained Wav2Vec2 model (~1-2GB). This may take a few minutes. ## Running the API ### Start the Flask Server ```bash cd myproject python …