Main Page

From ALIZE wiki
Jump to: navigation, search


The ALIZE Project

ALIZE Project - Open source platform (LGPL) for biometrics authentification

The main objective of this project is to facilitate access to biometric technologies for academic's (centre for research and education) and industrial's. To achieve this objective, ALIZE project provide an efficient, modular, and easy to master platform, capable of managing differents biometrics tasks, in different environments. One of the expected impact of the project is to animate a strong community of users of the ALIZE platform , composed of academic and industrial players. Major originality of ALIZE, is proposal of a single recognition engine for multiple modalities: primarily voice and face. ALIZE is also available on different operating systems(Linux / Windows / Mac OSX). For users of ALIZE, adoption of a unique and free recognition engine, allows to focus on research and development efforts on specific aspects of the modality studied, rather than on the development and monitoring of complex low level softwares engines.

With ALIZE, it's becomes easier to integrate new methods based on the know-how acquired over those already studied. Furthermore this approach simplifies the integration of biometrics in demonstrators or products, and fusion of different modalities. This is essential because multimodality should play a significant role in next generation of biometric systems. ALIZE is the continuation of previous biometrics projects TECHNOLANGUE AGILE/ALIZE completed in January 2005 and Mistral project (founded by French National Research Agency) completed in 2009.

Biometrics library ALIZE is today widely used by an international community of academic's and industrial's users. Those users are very active and prolific (as shown on the "project pages of ALIZE Website), and demonstrate that free software, when published trough a scientific collaboration, with respect to the rules of software engineering (UML standards), have great potential. ALIZE has obtained very good results in major international evaluation campaigns - NIST SRE, RT - and French - ESTER. The performance of ALIZE will also be tested by modality modality, through regular participation in various international campaigns.

ALIZE Project on the web


How to cite ALIZE

  • For reference articles and bibtex, please see this page.

ALIZE User guide


  • URL of the ALIZE repositories to access to the latest SVN versions

Interesting ALIZE Wiki pages

ALIZE for i-Vector

  • Statistics Computing statistics for JFA and TotalVariability estimation
  • TotalVariability Computing TotalVariability matrix
  • I-Vectors How to work in Total Variability space and demo
  • IvExtractor How to extract i-vectors
  • IvNorm How to normalize i-vectors
  • IvTest i-vector comparison using different types of scoring methods

Specialized ALIZE applications pages

Community around ALIZE

An open source community

ALIZE project is opened to external contributions.

Contribute to the ALIZE wiki

Every contribution on this wiki his highly welcomed ! Please consult the User's Guide for information on using the Wiki software used for ALIZE wiki.

Use ALIZE with Praat

Use some of the ALIZE tools as Praat Plugin.