Search

25 May 2004: κ is a scalingfactor and can be adjusted to improve test-set performance. ... Table 3. Comparison of WER for MPE trained HLDA and EM-LLT models on thedev01sub test set.

29 Apr 2004: There are some small gains using ensemble techniques, but the gains are notconsistent across the dev03f and eval03f test sets. ... The PFM combined with LDCDB98 SULMs can give some small gains, but thepatterns are not consistent across the dev03f and

27 May 2004: GI unadapted single pass decoding system. Table 2 gives WER on the test sets using the5xRT P1-P2system. ... thedev03 and eval03 test sets respectively.These results were even slightly better than the full 10xRT CU-.

25 Mar 2004: http://www.nist.gov/speech/tests/rt/rt2003/spring/docs/ref-cookbook-v2_4.pdf , 17th March 2003. ... 6] NIST, “Benchmark Tests : Rich Transcription (RT),”http://www.nist.gov/speech/tests/rt/. [7] S.E. Tranter, K.

23 Mar 2004: The warp-factors for test-data are estimated using the VTLN-MLE model.dev03 eval03. ... System Training Test EstimationAdaptation Adaptation MLE MPE. GI —— 33.4 30.4.

27 May 2004: The evaluation test set used in the 2003 DARPA/NIST RichTranscription evaluation (eval03) contains data from the LDC Fishercollection1 and from Switchboard II phase 5. ... Vocal TractLength Normalisation (VTLN) was used in training and test bywarping the

22 Mar 2004: SU-Detection System for RT-03F Evaluation. Results for the dev03 (3hrs) and eval03 (3hrs) test sets:.

28 Jul 2004: 5 Evaluation. To test our global localisation system, we constructed a database of views by photograph-ing all the buildings in the main shopping street in Cambridge’s city centre, as

22 May 2004: The training data set comprised 3990 utterances. All1200 test utterances (feb89, oct89, feb91, sep92), tst, and a 300utterance subset of the training data, trn, with a simple word-pair grammar, were

26 Jan 2004: The test set used wasthe subset of the 1997 Hub5 evaluation set used in [9]. ... The test set error rates for all the configurations describedabove are depicted in Table 3.

8 Oct 2004: 39-dimensional single and two observation noise component baseline FAHMMs. on the 1200 utterance test set and 300 utterance train set. ... Oracle: 1.18 (test), 0.08 (train). Idiot: 63.43 (test), 52.45 (train). 95.

11 Feb 2004: 5.2 Calibration Performance. The static performance of the HMD calibration methodused was evaluated by performing eight full stereo calibra-tions on the test model shown in Figure 6.

26 Nov 2004: Finally Section 4 contains the results of applying this technique on different test objects. ... 4 Results. Two ultrasound phantoms were used to test the attenuation estimation algorithms.

26 Jan 2004: The test data set was the. 12. Table 4: Order of number of free parameters and word error rates in Minitrain FAHMM systemwith k = 13.

13 May 2004: of shows inthe test set.

25 May 2004: Test sets iterations MLLR H-cri DLT MPE-DLTs3-dev 1 ite 13.2 12.4 12.2s3-eval 1 ite 11.1 10.3 10.1s3-dev 3 ite

18 Jun 2004: n}as test functions to measure deviations from the desired evolution properties (3)and (4). ... div(gN) ei. Using Green’s formula, we can. move one differentiation onto the test function ei and hence trade the secondorder derivatives for a boundary

29 Jan 2004: For the evaluation, 1200test utterances (feb89, oct89, feb91, sep92), test, and a randomly selected 300 utterance subsetof the training data, train, were used. ... Thisdoes not seem to be the case in the two component systems where the test set

29 May 2004: The choice of Dj affects the cri-terion convergence and test set generalization [3, 4].