Music Genre Classification and Song Similarity Search

Overview

This project involves the classification of music genres using audio data and the calculation of similarity between songs based on their audio features. The steps include:

1. Feature extraction from audio files.
2. Scaling the features.
3. Computing the cosine similarity between songs.
4. Finding the most similar songs to a given song using the computed similarity matrix.
5. Playing the selected song's audio.

Dependencies

Before the working project, update general_path part accoring to your workplace.( Second cell in MusicGenreClassification.ipynb)

The following libraries are required for this project:

• pandas: For data manipulation and creating DataFrame objects.
• scikit-learn: For scaling the features and calculating cosine similarity.
• IPython: For audio playback in Jupyter notebooks.
• librosa: For audio feature extraction (e.g., MFCC, Chroma).
• numpy: For numerical operations.

pip install pandas scikit-learn librosa numpy IPython