Search

27 May 2004: INTRODUCTION. The majority of state-of-the-art speech recognition systems aretrained on found data, for example broadcast news and telephoneconversations. ... This is consistent with the gains that were obtainedon the Broadcast News task [11].

23 Mar 2004: New data sources becoming available for RT04. • Finding out how to make most effective use of new data. • ... Also investigating many new algorithms. • Collaboration is working very well across teams: workshops, data sharing etc. •

27 May 2004: The large rescoring lattices produced in the first stage are thenrescored with the adapted triphones and new lattices are gener-ated on output. ... Niesler, A.Tuerk, E.W.D. Whittaker and S.J. Young, “The 1997 HTKBroadcast News Transcription System”

29 Apr 2004: Reference: The LIMSI Broadcast News transcription systemJean-Luc Gauvain, Lori Lamel and Gilles AddaSpeech Communication, Volume 37, Issues 1-2, May 2002, Pages 89-108.

26 Jan 2004: The new observation matrix has to be estimated row by row asin SFA [3]. ... The lth row vector ĉl of the new observation matrix, Ĉj ,can be written as.

25 May 2004: Select manageable subset as baseline for investigation of new techniques400h, balanced for gender, line conditions, topics. • ... WER on eval03 and dev04, unadapted, 2003 trigram. • New Fisher 400h set gives very similar performance to old 520h one.

18 Jun 2004: The majority of state-of-the-art speech recognition systems aretrained onfounddata, for example broadcast news and telephoneconversations. ... This is consistent with the gains that were obtainedon the Broadcast News task [11].

23 Mar 2004: Woodland et al.: Cambridge STT Overview. Outline. • Broadcast News. • Lightly supervised discriminative training. • ... Broadcast News: Segmentation and Clustering. • Running our evaluation system on different segmentations and clusters. •

25 May 2004: NorthAmerican Business News Spoke 3 task.

28 Jul 2004: 14] Z. Zhang. A flexible new technique for camera calibration. IEEE Transactions on Pattern Analysis and Machine Intelli-gence, 22(11):1330–1334, 2000.