Lyon's auditory model for Python

Python port of Lyon's model calculation from Auditory Toolbox. Original version of Auditory Toolbox is written in C and MATLAB by Malcolm Slaney.

Package contents

• Modified soscascade.c, agc.c and sosfilters.c (removed MEX-related part).
• ctypes wrapper for soscascade(), agc() and sosfilters() calls.
• Translation from MATLAB to Python for files necessary for successfull call to LyonPassiveEar().

Dependecies

If you plan to build manually and run tests, you'll need libcheck installed. On Ubuntu 18.10 run:

sudo apt-get install check

Installation

PIP

pip install lyon

Manual

Build a library and copy to installation location:

make -C c_src lib
cp c_src/liblyon.so lyon/

you can verify that library is properly built by running a test suite:

make -C c_src test

Usage

The following code computes cochleogram for a sample sound:

from lyon import LyonCalc

calc = LyonCalc()
waveform, sample_rate = load('audio/file/path.wav')
decimation_factor = 64
coch = calc.lyon_passive_ear(waveform, sample_rate, decimation_factor)
print(coch.shape)

The code above should output shape of resulting auditory nerve response: [<number of samples / decimation_factor>, 86].

See examples/lyon_examples.py for examples on running other functions.

Links

• AuditoryToolbox page
• Report describing function in toolbox