1 / 16

Music Analysis

This research project focuses on analyzing audio data to identify music genres using statistical, signal processing, and machine learning methods. It explores techniques such as Fourier Transform for spectral analysis and fractal dimension measurements for genre identification. By employing feed-forward neural networks, the study examines the complexity and structure of music, distinguishing between genres via various spectral aggregations. Results indicate a correlation between frequency and magnitude properties, providing insights into effective music classification methods and future avenues for research in machine learning and signal processing.

lysa
Download Presentation

Music Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Josiah Boning TJHSST Senior Research Project Computer Systems Lab, 2007-2008 Music Analysis

  2. Purpose • Framework for study of audio data • Components: • Statistical • Signal Processing • Machine Learning • Uses: • Aircraft identification • Speech recognition and synthesis • And more... • Ideal: computer recognition of music

  3. Background • Bigarelle and Iost (1999)‏ • Music genre can be identified by fractal dimension • Basilie et al. (2004)‏ • Music genre can be identified by machine learning algorithms • Used discrete MIDI data

  4. Methods • Data Processing • Spectral Analysis: Fourier Transform • Fractal Dimension: Variation and ANAM Methods • Machine Learning • Feed-Forward Neural Network

  5. Time domain to frequency domain Spectral Analysis – Fourier Transform

  6. Fourier Transform

  7. Spectrogram • Many Fourier transforms sequentially

  8. Frequency Aggregate • Horizontal sum of spectrogram

  9. Magnitude Aggregate • Vertical sum of spectrogram • Can perform second Fourier transform

  10. Fractal Dimension • Bigerelle and Iost – used to distinguish genre • Variation Method: • ANAM Method:

  11. Fractal Dimension • Inaccurate calculations • Correct values around 1.6-1.9 • Variation: ~1.16 • ANAM: ~2.25 • Interpolation between samples didn’t help

  12. Machine Learning • Neural networks • Feed-Forward

  13. Neurons • Each node performs weighted, rescaled sum of input values • Scaling: Sigmoid function

  14. Results • Frequency aggregate – songs are similar!

  15. Results • Magnitude and frequency: negative correlation • Not shared bywhite noise • Fourier transform of magnitude aggregate • Meaningless

  16. Where next? • Move code from driver to Fourier module • Otherwise, well organized • Fix fractal dimension calculations • Neural network learning algorithms • Beat detection • Other signal processing techniques

More Related