README

Python packages for audio onset detection Evaluation

Structure

In the "src" dir the tool is organised into python packages as followed:-

Eval : statistical tools for evaluating onsets prediction vs ground-truth
Predict : global package to generate pool and onsets text files
- OnsetNovelty : package grouping multiple onset novelty functions
- Slice : package grouping multiple multidimensional "peak-picking" algorithms
Utils : common routines and functions for dealing with files and the environment.

conf.py at the top level holds environment specific things - customise this for your setup.

Specify folders in conf.py

for already computed onset textfiles : run Eval/__init__.py

for computing : specify desired methods in OnsetNovelty's and Slice's __init__.py then run Predict's init.py

you can use last computation using pool cache (basicaly set fromFile=true in config.py)

http://pydev.org/manual_101_install.html
You may need to install Oracle Java (http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html)
Then set Java -> Installed JREs -> to Java SE 8
Restart and in Preferences you should now have PyDev appearing
Set Interpreters > Python Interpreter to "/usr/local/bin/python" or whatever

Make sure to set the "src" directory as the PYTHONPATH to detect the packages.

*opencv