Pronunciation Trainer API - OpenAI Version

A cloud-based pronunciation analysis API using OpenAI's Whisper and TTS services.

Features

• Audio transcription using OpenAI Whisper API
• Pronunciation scoring and analysis
• IPA (International Phonetic Alphabet) transcription
• Text-to-speech using OpenAI TTS
• Firebase integration for data storage
• Statistics and analytics
• Fast and scalable

Setup

1. Install dependencies:

python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

2. Configure environment variables in .env

3. Add Firebase credentials file

4. Run the server:

uvicorn app.main:app --reload

API Endpoints

Pronunciation Analysis

• POST /api/transcribe - Transcribe audio file
• POST /api/analyze - Full pronunciation analysis
• POST /api/grade - Get pronunciation grade
• POST /api/ipa - Get IPA transcription

Text-to-Speech

• POST /api/tts - Generate pronunciation audio
• POST /api/tts/slow - Generate slow pronunciation

Data Management

• GET /api/recordings - List all recordings
• GET /api/recordings/{id} - Get specific recording
• DELETE /api/recordings/{id} - Delete recording
• GET /api/statistics - Get usage statistics

Utility

• GET /api/health - Health check
• GET /api/phrases - Get practice phrases