NIST SPONSORS 2000 NIST SPEAKER RECOGNITION EVALUATION WORKSHOP.NIST (National Institute of Standards & Technology, Washington, DC, www.nist.gov) The standards-defining agency of the U.S. government, formerly the National Bureau of Standards. It is one of three agencies that fall under the Technology Administration (www.technology. organized and hosted the 2000 NIST Speaker Recognition Evaluation Workshop in June June: see month. 2000, in Linthicum, MD. The purpose of the workshop was to review performance of systems in the evaluation, to discuss trends in text-independent speaker recognition, and to plan the next evaluation. The participating and contributing sites were from industry, academia, and governments in Australia, France, India, Israel, South Africa South Africa, Afrikaans Suid-Afrika, officially Republic of South Africa, republic (2005 est. pop. 44,344,000), 471,442 sq mi (1,221,037 sq km), S Africa. , Spain, and the United States United States, officially United States of America, republic (2005 est. pop. 295,734,000), 3,539,227 sq mi (9,166,598 sq km), North America. The United States is the world's third largest country in population and the fourth largest country in area. . For the past 5 years, NIST has organized evaluations of text-independent speaker recognition performance on conversational telephone speech. The tasks included one-speaker detection, two-speaker detection, speaker tracking, and speaker segmentation. The 2000 evaluation included for the first time some data in Castilian Spanish Castilian Spanish refers to some dialects of the Spanish language as spoken in Spain, also known as Spanish Spanish or Spanish from Spain. Although castellano and other non-English languages. These evaluations are open to all interested sites. A workshop is held following each evaluation for participants to review evaluation performance and the state of speaker recognition technology and to consider plans for future evaluations. These evaluations have become internationally recognized as the leading measure of state-of-the-art performance in text-independent speaker recognition. |
|
||||||||||||||||||

Printer friendly
Cite/link
Email
Feedback
Reader Opinion