API descriptionΒΆ

Copyright 2014-2017 Anthony Larcher and Sylvain Meignier

Authors:Anthony LACHER, Sylvain MEIGNIER & Kong Aik LEE
Version:1.1.6 of 2016/10/31
This package is the core of the SIDEKIT toolkit.
While developing SIDEKIT, we tried to keep in mind two

1. limit the number of classes to allow better readability
2. make SIDEKIT compatible with existing tools

To reach this target, we have created four Main Classes
(FeaturesExtractor`, FeaturesServer, Mixture and StatServer) which can be used
together with a number of tools available in companion
modules (sidekit_io and sv_utils).

Front-end and back-end processing such as acoustic feature extraction
and score analysis are handled in two packages: frontend and bosaris.
The frontend package include a number of tools to extract and normalize
the acoustic features as well as detecting the high energy frames for voice
activity detection. The bosaris package consists of the translation of
a part of the BOSARIS toolkit available on this webpage.
The current python implementation of the BOSARIS toolkit does not include tools
for calibration and fusion but only the core structures that are used to manage
enrollment lists, trial definitions and scores.
The authors would like to thank Niko Brummer and AGNITIO to allow them to distribute
this version of the BOSARIS toolkit.


The bosaris package which is released together with SIDEKIT is distributed under a different license. The intellectual property belongs to the original authors of the toolkit.