NIST 2006 Speaker Recognition Evaluation - Participants

Introduction

The table below lists the participants in this year's speaker recognition evaluation.

The table below lists the submissions, the status of the submissions, and the tasks which the sites participated in. Each row represents a different system. If multiple submissions for a given test were received, the one site indicated as primary is marked with a 'P'. Single submission for a given test is marked as primary also regardless if it came from the best system or not. No adaptation is marked with 'N', unsupervised adaptation is marked as 'U', and 2005 system is marked as '2005'

As required by the evaluation plan, sites were to submit a brief description of the system (the algorithms) used to produce their results. This is to include the CPU execution time that was required to proceess the test data as if the test were run on a single CPU. Sites were to describe the CPU used and the amount of memory used. The system descriptions are also linked in the table.

Please note that sites are free to do what they wish with their own results and system descriptions, but can neither redistribute nor publish results from other sites without the site's expressed permission.

The asterisk (*) in the On Time column below indicates that the submission is a bug-fixed. A bug-fix means there is a change in the system's algorithm, however, minor. Output format errors do not count as bug-fixes.