Light

ctralie / graphditty Goto Github PK

View Code? Open in Web Editor NEW

38.0 5.0 7.0 23.95 MB

Similarity network fusion of features for clean self-similarity matrices in audio, as well as Javascript tools to play SSMs and spring graphs synchronized to audio

License: Apache License 2.0

Python 31.87% HTML 1.92% JavaScript 24.03% CSS 0.78% TeX 41.38% Shell 0.03%

graphditty's Introduction

GraphDitty

The purpose of this code is to compute clean weighted adjacency matrices for audio files using similarity network fusion, which can be used to do clean structure analysis. This is based on ideas from the following papers:

Christopher J Tralie and Brian McFee. "Enhanced Hierarchical Music Structure Annotations via Feature Level Similarity Fusion. International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2019.
Christopher J Tralie. GraphDitty: A Software Suite for Geometric Music Structure Visualization. In 19th International Society for Music Information Retrieval (ISMIR), 2018, Late Breaking Demo Session.
Christopher J Tralie. ``MFCC and HPCP fusion for robust cover song identification.'' In 18th International Society for Music Information Retrieval (ISMIR), 2017.

Example Fused Adjacency Matrix on Michael Jacksons's Bad

Dependencies

numpy/scipy/matplotlib
librosa

Running

To see all options, run the script as follows

python SongStructure.py --help

There are options to view each iteration of the fusion. By default, the results are saved to a .mat file called ``out.mat'' and to a file "out.json" (you can changes this via command line options). For example, the above figure was generated with the following call

python SongStructure.py --filename MJ.mp3 --jsonfilename MJ.json

You can open the .json file in the web page "Viewer/index.html" to interactively view the similarity matrix and Laplacian eigenvectors, as well as as force graph and diffusion maps based on the similarity matrix.

graphditty's People

Contributors

Stargazers

Watchers

Forkers

bmcfee atine ursinus-cs476-f2020 octophi silvadirceu knut0815 florianthalmann

graphditty's Issues

Dependencies

Hi, can you list down the exact versions of your dependencies?
in particular,

python2.? or python3.?
librosa==?
scipy==?

I am using python 3.6, and at one of the later versions of librosa==0.7.0, it is required for scipy>=1.0.0.

When i ran the program, i met into the following error

Traceback (most recent call last):
  File "SongStructure.py", line 300, in <module>
    saveResultsJSON(opt.filename, res['times'], res['Ws'], opt.neigs, opt.jsonfilename, opt.diffusion_znormalize)
  File "/data3/lootiangkuan/AudioHashing/QbH/v2_GraphDitty/SongStructureGUI.py", line 151, in saveResultsJSON
    Results['W'] = getBase64PNGImage(WOut, 'magma_r', logfloor_quantile=0.01)
  File "/data3/lootiangkuan/AudioHashing/QbH/v2_GraphDitty/SongStructureGUI.py", line 58, in getBase64PNGImage
    scipy.misc.imsave("temp.png", C)
AttributeError: module 'scipy.misc' has no attribute 'imsave'

According to this question , it seems i require scipy==0.17.0

librosa.feature.melspectrogram function type error

very small typo

https://github.com/ctralie/GraphDitty/blob/master/SongStructure.py#L168

    S = librosa.feature.melspectrogram(y, sr=sr, n_mels=128, hop_length=hop_length)

throws
TypeError: melspectrogram() takes 0 positional arguments but 1 positional argument (and 2 keyword-only arguments) were given
fix:

    S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=128, hop_length=hop_length)

thank you for making this code available, your papers were very insightful.

what does negative values in SSMs mean?

I calculated 'Fused', 'MFCCs', 'Chromas' and 'Tempogram' matrices for a song, the result shows some of the SSMs has negative values, specifically, their minimums and maximums are:

Fused: 0.0 ~ 3.159
MFCCs: -0.054 ~ 1.000
Chromas: -0.105 ~ 1.000
Tempogram: 0.0 ~ 1.000

I wonder what the negative values mean in SSM, as i thougt the self-similarity metrics should be a positive value, ideally in range [0,1]. So what does negative values in SSMs mean, or is that a bug?

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.