🎵 PyShazam: Audio Fingerprinting & Identification

📌 Project Overview

PyShazam is a desktop application that identifies songs from audio clips, similar to Shazam. It uses Digital Signal Processing (DSP) techniques to extract unique fingerprints from audio files (Spectrogram Peaks) and matches them against a pre-built database.

The tool also features a Spectrogram Visualizer and an Audio Mixer to test the system's robustness against noise and overlapping tracks.

⚙️ How it Works (The Algorithm)

1. Spectrogram Generation: Converts audio signals to the frequency domain using Short-Time Fourier Transform (STFT) via Librosa.
2. Feature Extraction: Identifies local maxima (peaks) in the spectrogram (Constellation Map).
3. Fingerprinting: Hashes the peaks based on their frequencies and time deltas to create a unique signature for each song.
4. Matching: Compares the input clip's hashes with the database (JSON) to find the highest similarity score.

🚀 Key Features

• Song Identification: Identifies tracks even from short or noisy clips.
• Similarity Score: Displays a confidence percentage for the matched song.
• Spectrogram Viewer: Visualizes the intensity of frequencies over time.
• Database Management: Automatically hashes and stores new songs in fingerprints.json.
• Audio Mixing: Ability to merge two distinct tracks to test the identification capability under interference.

🛠️ Tech Stack

• Core Logic: Python, NumPy, Pandas.
• Audio Processing: Librosa (STFT, Mel-Spectrogram).
• GUI: PyQt5.
• Storage: JSON (Lightweight Hash Database).

🚀 Installation & Usage

1. Clone the Repo:

   git clone [https://github.com/mariamashraf731/PyShazam-Audio-Fingerprinter.git](https://github.com/mariamashraf731/PyShazam-Audio-Fingerprinter.git)

2. Install Dependencies:

   pip install -r requirements.txt

3. Run the App:

   python src/main.py