Hans-Werner Fink, Heinz Schmid, et al.
Journal of the Optical Society of America A: Optics and Image Science, and Vision
This paper describes the IBM approach to Broadcast News (BN) transcription. Typical problems in the BN transcription task are segmentation, clustering, acoustic modeling, language modeling and acoustic model adaptation. This paper presents new algorithms for each of these focus problems. Some key ideas include Bayesian information criterion (BIC) (for segmentation, clustering and acoustic modeling) and speaker/cluster adapted training (SAT/CAT). © 2002 Elsevier Science B.V. All rights reserved.
Hans-Werner Fink, Heinz Schmid, et al.
Journal of the Optical Society of America A: Optics and Image Science, and Vision
Jianchang Mao, Patrick J. Flynn, et al.
Computer Vision and Image Understanding
Dorit Nuzman, David Maze, et al.
SYSTOR 2011
Lalit R Bahl, Steven V. De Gennaro, et al.
IEEE Transactions on Speech and Audio Processing