Skip to content

Latest commit

 

History

History
23 lines (18 loc) · 1.3 KB

README.md

File metadata and controls

23 lines (18 loc) · 1.3 KB
Raw

Genere Musicale Recognition with K-Nearest Neighbors (KNN) Algorithm

Overview

This project explores an approach based on the use of MATLAB and the K-Nearest Neighbors (KNN) algorithm to classify electronic music genres, specifically ambient minimal house. Spectral and temporal features are extracted from the music audio files for classification.

Features in Time Domain

  • Energy: The magnitude of the signal.
  • Zero Crossing Rate: The frequency at which the signal changes from positive to negative or back.

Features in Frequency Domain

  • Spectral Centroid: The main point of the spectrum distribution.
  • Spectral Spread: The standard deviation of the spectrum distribution.
  • Spectral Rolloff: The amount of energy accumulated until a certain point in the frequency.
  • Mel-Frequency Cepstral Coefficients (MFCC): Represents the spectrum bands according to the mel-scale, an isophonic (mostly subjective) coefficient.

Usage

  1. Ensure you have MATLAB installed on your system.
  2. Download the repository.
  3. Open and run the MATLAB script The_main.m to train the KNN model and classify the music genres.
  4. View the results and evaluate the performance of the KNN algorithm.

Credits

This project was developed by Alessandro Scalambrino