README.md

metadata

title: Aura Emotion Detection API
emoji: 🎤
colorFrom: blue
colorTo: purple
sdk: docker
pinned: false

🎤 Aura Emotion Detection API

Real-time emotion detection from audio using Wav2Vec2 model from Hugging Face.

🚀 Features

• Real-time Emotion Detection: Uses superb/wav2vec2-base-superb-er model
• Multiple Audio Formats: Supports WAV, MP3, WebM, and more
• Fast Processing: Optimized for real-time analysis
• REST API: Easy integration with any frontend

📖 API Endpoints

Health Check

GET /health

Predict Emotion

POST /predict
Content-Type: multipart/form-data
Body: audio file (WAV, MP3, WebM, etc.)

Response:

{
  "emotion": "happy",
  "confidence": 0.85,
  "model": "Wav2Vec2 (Hugging Face)"
}

🎯 Supported Emotions

• happy - Joyful, cheerful
• sad - Sad, melancholic
• angry - Angry, frustrated
• calm - Calm, relaxed
• excited - Excited, energetic
• neutral - Neutral, no strong emotion

🛠️ Technology Stack

• Framework: FastAPI
• Model: Wav2Vec2 (superb/wav2vec2-base-superb-er)
• Audio Processing: librosa, soundfile, pydub
• ML Framework: PyTorch, Hugging Face Transformers

📝 Usage Example

import requests

# Upload audio file
with open('audio.wav', 'rb') as f:
    files = {'audio': f}
    response = requests.post(
        'https://your-username-aura-emotion-api.hf.space/predict',
        files=files
    )
    
result = response.json()
print(f"Detected emotion: {result['emotion']}")

🌐 Frontend Integration

The frontend is deployed on Vercel and connects to this API for real-time emotion detection from microphone input.

📝 License

MIT License