Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

i-vector Machine Learning Challenge

  1. Speaker Recognition i-Vector Machine Learning Challenge

NIST began coordinating a special i-vector challenge in late 2013 based on data used in previous NIST Speaker Recognition Evaluations (SREs), and early results were featured in a special session of the Odyssey 2014 International Workshop on Speaker and Language Recognition. The official evaluation period has ended, however the platform will still score system submissions.

More information regarding the challenge can be found in the evaluation plan.
  1. Language Recognition i-Vector Machine Learning Challenge 

NIST will coordinate a special i-vector challenge beginning in mid-2015 based on data used in previous NIST Language Recognition Evaluations (LREs). This challenge focuses on the development of new methods for using i-vectors for language recognition in the context of conversational telephone or narrowband broadcast speech. 

It is designed to foster research progress, including goals of: 

  • Exploring new ideas in machine learning for use in language recognition
  • Making the language recognition field accessible to more participants from the machine learning community
  • Improving the performance of language recognition technology.

More information regarding the challenge can be found in the evaluation plan.

In order to foster interest from the Machine Learning Community in the fields of Speaker Recognition and Language Recognition, the National Institute of Standards and Technology (NIST) began organizing evaluations in the format of challenges. The challenges are based on the i-vector paradigm widely used by state-of-the-art speaker and language recognition systems. By providing such i-vectors directly and not utilizing audio data, the evaluation is intended to be readily accessible to participants from outside the audio processing field.

In order to register and participate, please visit the challenge site: 

Updates will be posted to this webpage. 

Created November 18, 2013, Updated August 25, 2016