Speaker Recognition: Building the Mixer 4 and 5 Corpora.
作者： Christopher Cieri , Kevin Walker , David Graff , Linda Brandschain , Abby Neely
DOI:
关键词:
摘要: The original Mixer corpus was designed to satisfy developing commercial and forensic needs. resulting corpora, Phases 1 through 5, have evolved support increasing variety of research tasks, including multilingual cross-channel recognition. 4 5 corpora feature a wider channels greater variation in the situations under which speech is recorded. This paper focuses on plans, progress results 5.
参考文章(3)
Christopher Cieri, Kevin Walker, David Graff, Linda Corson, Resources for new research directions in speaker recognition: the mixer 3, 4 and 5 corpora. conference of the international speech communication association. pp. 950- 953 ,(2007)
Mark Liberman, Christopher Cieri, Mark A. Przybocki, Walter D. Andrews, Shudong Huang, Kevin Walker, Alvin F. Martin, Joseph P. Campbell, John J. Godfrey, George R. Doddington, Hirotaka Nakasone, The Mixer and Transcript Reading Corpora: Resources for Multilingual, Crosschannel Speaker Recognition Research language resources and evaluation. pp. 117- 120 ,(2006)
Christopher Cieri, Joseph P. Campbell, Hirotaka Nakasone, Kevin Walker, David Miller, The Mixer Corpus of Multilingual, Multichannel Speaker Recognition Data language resources and evaluation. ,(2004) , 10.21236/ADA523534