Shazam-like Song Identifier

A Python desktop application that identifies songs and enables users to mix songs with weighted averages. It provides advanced analysis and matching features using mel-spectrogram, MFCC, and chroma-STFT, with perceptual hashing for efficient comparison.

---

Features

• Song Visualization:
  • Double-click on a graph to load a song.
• Song Mixing:
  • Use a slider to add a weighted average between two songs.
• Song Matching:
  • A table on the left displays top matches for the uploaded or mixed song.
• Media Player:
  • Small players for each song with play, pause, reset, and seek functionalities.
• Custom Database:
  • Users can build a custom database by running FeatureExtraction.py after adding their folder of music.

---

Installation

1. Clone the repository:

   https://github.com/Ahmed-Hajhamed/Shazam-like-Song-Identifier.git
   

2. Install dependencies:

   pip install -r requirements.txt

Usage

-Run the application: bash python app.py

• Load songs by double-clicking the graphs.
• Mix songs using the slider to adjust the weighted average.
• View matching results in the table.
• Use the media players to listen to and seek through the songs.
• To create a custom database of music:
  1. Add your folder of music to bash process_folder("path_to_folder") function in FeatureExtraction.py. Make sure all audio files are in the subfolder.
  2. Run bash FeatureExtraction.py to create the database and you are ready to go!

Requirements Python: Version 3.8 or higher Libraries: librosa PyQt5 numpy pyqtgraph qt_material ImageHash pandas

Screenshots

Load Songs By Double-Clicking on Graphs

Adjust The Weight of Each Song

Interactive Media Player

Contributing

Contributions are welcome! Please open an issue or submit a pull request if you have suggestions or bug fixes.

Contact

For questions or support, please reach out via the GitHub issues page or email: [email protected].