| Document Date: 2006-11-14 12:56:40
Open Document File Size: 976,76 KBShare Result on Facebook
Company IBM / / Country United States / / Currency USD / / / Event Reorganization / / Facility University of Washington / Building Large-Context Decoding Graphs We / / IndustryTerm k-means algorithm / English language telephone conversations / conversational telephony speech / search optimization strategies / bank / search errors / telephone-quality speech / search speed-ups / technology / weighted outer products / telephone conversations / diagonal-covariance systems / real time / search graph / large-context search space representation / commodity hardware / outer product / speech recognition technology / collection protocols / power law / basic decoding algorithms / consensus processing / conversational telephony test data / technology threads / representative systems / token passing algorithm / technology improvements / speaker-independent and speaker-adapted systems / search space / / MarketIndex set 500 / / Movie SPLICE / / MusicAlbum SPEECH / / OperatingSystem Linux / / Organization University of Washington / Defense Advanced Research Projects Agency / FSA / / Person Daniel Povey / Geoffrey Zweig / Timothy J. Hazen / George Saon / Brian Kingsbury / Hagen Soltau / Lidia Mangu / / Position standard speaker / speaker / real time RT / RT System Architecture RT / RT / associate editor / Fisher / representative / / Product B-2 / / Technology basic decoding algorithms / decoding algorithms / speech recognition / k-means algorithm / finite state machine / Linux / speech recognition technology / The algorithm / token passing algorithm / Digital Object Identifier / collection protocols / / URL http /
SocialTag |
1596 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 14, NO. 5, SEPTEMBER 2006 Advances in Speech Transcription at IBM Under the DARPA EARS Program