Processing mp3 by Python3 in Linux

 

 

$ sudo apt install llvm-8
 

$ locate llvm-config

$ LLVM_CONFIG=/usr/bin/llvm-config-8 pip3 install librosa --user
 

 

packages: llvmlite, numba, soundfile, resampy, audioread, librosa

 

PCA/WPCA

Video Lectures
[1] SVD to PCA https://www.youtube.com/watch?v=mBcLRGuAFUk

Lectures
[1] https://skymind.ai/wiki/eigenvector
[2] https://www.projectrhea.org/rhea/index.php/PCA_Theory_Examples


[3] PCA Theory Example http://www.cs.otago.ac.nz/cosc453/student_tutorials/principal_components.pdf

[4] Animation 2D PCA https://www.intmath.com/matrices-determinants/eigenvalues-eigenvectors-concept-applet.php


[5] PCA on OpenCV
https://robospace.wordpress.com/2013/10/09/object-orientation-principal-component-analysis-opencv/


[6] Wavelets, Fourier, Kalman http://www.cs.otago.ac.nz/cosc453/





Resources:
[1] https://machinelearningmastery.com/calculate-principal-component-analysis-scratch-python/
[2] https://jakevdp.github.io/PythonDataScienceHandbook/05.09-principal-component-analysis.html
[3] Draw 2D principal axis PCA https://stackoverflow.com/questions/18299523/basic-example-for-pca-with-matplotlib

Audio Feature Extraction

https://github.com/tuwien-musicir/rp_extract
https://github.com/tyiannak/pyAudioAnalysis

Machine Learning topics 2019

References:
[1] Encoding categorical data https://medium.com/@rabinpoudyal1995/encoding-categorical-data-in-python-fab150d6e21b
[2] Espectral clustering https://towardsdatascience.com/spectral-clustering-82d3cff3d3b7

Signal Processing - Sound Processing

 Resources:
[1] Course and Tutorial https://github.com/AllenDowney/ThinkDSP

Lecture about SoundScape

Acoustic scenes, Acoustic events and Feature aggregation

D-CASE2013

Environment sound recognition using short-time feature aggregation
https://link.springer.com/article/10.1007/s10844-017-0481-4

Task of segmentation
  Speech recognition
  Musical information retrieval(MIR)
  Environment sound
  Survillance
  Index of video content
Task of acoustic events classification
  Statistic large set of features and K-NN
  Feature agregation (Bag of feaures)
  HMM/GMM(better for segementation and annotation for small scales)
  Deep learning
  Blind segmentation and SVM


Recurrence quantification analysis(RQA), RQA Features
Short Time Fourier Transform (STFT)