Faceify

Names: Anvita Yerramsetty, Vije Kirubanandan, Marina Chen, Leander Stephen https://github.com/lestep1225118/hackathon-project.git

This project uses computer vision and machine learning to analyze facial expressions in images and automatically generate personalized Spotify playlists that match the detected emotions.

Features

• Face Detection: Uses OpenCV to detect and extract faces from images

• Emotion Analysis: Employs DeepFace to analyze facial expressions and detect emotions:

  • Primary emotions: Happy, Sad, Angry, Neutral, Surprised, etc.
  • Secondary emotions through emotion blending
  • Confidence scoring for emotion detection

• Spotify Integration:

  • Creates custom playlists based on detected emotions
  • Uses emotion-specific song seeds and audio features
  • Automatically sets playlist cover image using the analyzed face
  • Generates diverse song selections through multiple search strategies

• Emotion-Music Mapping:

  • Maps detected emotions to musical characteristics
  • Considers valence, energy, tempo, and danceability
  • Blends multiple emotions for more nuanced playlist generation
  • Includes fallback options for uncertain emotion detection

Technologies Used

• Python 3.6+
• Flask (Web Framework)
• OpenCV (Computer Vision)
• DeepFace (Emotion Detection)
• Spotipy (Spotify API client)
• PIL (Python Imaging Library)
• NumPy (Numerical Computing)
• Werkzeug (WSGI Utilities)
• python-dotenv (Environment Management)

Setup

1. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

2. Install dependencies:

   pip install -r requirements.txt

3. Set up environment variables:

   • Create a .env file in the project root
   • Add your Spotify API credentials:

     SPOTIFY_CLIENT_ID=your_client_id
     SPOTIFY_CLIENT_SECRET=your_client_secret
     SPOTIFY_REDIRECT_URI=http://localhost:8000/callback
     

4. Run the app:

   python app/main.py

Usage

1. Run the program
2. Enter the path to an image containing a face
3. The program will:
   • Detect and analyze the face
   • Show detailed emotion analysis
   • Create a custom Spotify playlist
   • Display the playlist URL
4. Type 'q' to quit the program

Technical Details

• Face detection using OpenCV's Haar Cascade Classifier
• Emotion detection using DeepFace with RetinaFace backend
• Flask web server for handling requests
• Spotify Web API for playlist creation and management
• Custom emotion mapping and song selection algorithms
• Support for multiple image formats (jpg, jpeg, png)

Requirements

See requirements.txt for complete list of dependencies