Search

16 Jul 2009: Cambridge UniversityEngineering Department. JHU Workshop 2009 20. Sequence Kernels for Speaker and Speech Recognition. ... 20, pp. 210–229, 2005. [3] H. Lodhi, C. Saunders, J. Shawe-Taylor, N.

7 Oct 2009: MPE PLP Original 20.5 22.8PLPMLP1350hr Fast 17.7 20.0. Semi-fast 17.7 19.9Rebuild 17.7 19.9. ... SystemWER. dev07 eval07 dev08. ML. PLP 21.1 22.9 25.1MLP1350hr 20.2 21.2 22.6PLPMLP200hr 18.7 20.1 21.6PLPMLP1350hr 18.6 19.8 21.2PLPMLP1350hr big

4 Aug 2009: 15. 20. 25. Lateral, mm. ISNR = 2.04 dB (log: 1.32 dB). ... 15. 20. 25. Lateral, mm. ISNR = 2.40 dB (log: 1.38 dB).

28 Jul 2009: and subsequently combinetheir output, by either switching between them [2] or byprobabilistically merging them [8, 15, 18, 20]. ... 20] P. Pérez, J. Vermaak, and A. Blake. Data fusion for visual track-ing with particles.

2 Mar 2009: 0.1 0.2 0.5 1 2 5 10 20 40. 0.1. 0.2. ... 0.5. 1. 2. 5. 10. 20. 40. False Alarm probability (in %). Mis. s pr. obab. ility. (in. %). GMMLLR. 128. λ128. λ.

28 Jan 2009: 1. depends partly on the size and shape of the compressor [20, 23]. ... The lateral and elevationalfocal depths are both 20 mm. 3.1 Mechanical stress artefacts.

2 Mar 2009: 9.15 10.38 11.11Approx 12.21 10.20 9.25 10.29.

20 Aug 2009: 4. Using normalised mutualinformation, the Nelder-Mead simplex algorithm converged in all cases out of which 80% requiredno restarts and 20% required one restart. ... Ultrasound in Medicine and Biology, 20(4):923–936, 1994. [7] D. F. Leotta, P.

28 Apr 2009: s)d = Adb. (s)ml bd. Cambridge UniversityEngineering Department. Tsinghua University April 2009 20. ... Rep. TR611, February2009, available from http://mi.eng.cam.ac.uk/mjfg/. [20] R. Gopinath, B. Ramabhadran, and S.

4 Aug 2009: Numerous. methods have been proposed [19, 20, 21, 22]. They differ from one another in how they estimate. ... 0.5. 1. 1.5. Err. or,. %. Lateral focus, mm. 9.20. mm.

2 Mar 2009: Initially, combination using aminEER. 3The results presented here differ from those previously reported in[20] and [31]. ... 0.1 0.2 0.5 1 2 5 10 20 40. 0.1. 0.2.

9 Mar 2009: 6 3 ANALYSIS. 20 40 60 80 100sample number. r2=a. r2=24 a. ... White noise was added toreduce the SNR to 20 dB. The raw strain image in Fig.

2 Mar 2009: As in iterativeMLLR [20], multiple iterations can be used to refine theestimation of MLLR. ... 10 15 20 25 30 35 40 45 50 550.5. 1.

18 Feb 2009: 2.2 Constrained MLLR. A tied mean and covariance transform can be applied to both mean vectors and covariance ma-trices, this is referred to as a constrained linear transform [20, ... µ(m)o = µ(m)s µ. (rm)b (20). Σ(m)o = Σ(m)s Σ. (rm)b (21).

28 Jul 2009: and subsequently combinetheir output, by either switching between them [2] or byprobabilistically merging them [8, 15, 18, 20]. ... 20] P. Pérez, J. Vermaak, and A. Blake. Data fusion for visual track-ing with particles.

28 Jul 2009: 2008), doi:10.1016/j.patrec.2008.04.005. and timing each stroke, we were able to estimate the hand labelingtime for one frame to be around 20–25 min. ... 5.35%. 26.41%. 3.34%. 66.52%. 33.07%. 47.97%. 86.90%. 81.01%. 95.90%. 0% 10% 20% 30% 40% 50% 60% 70

28 Jul 2009: and subsequently combinetheir output, by either switching between them [2] or byprobabilistically merging them [8, 15, 18, 20]. ... 20] P. Pérez, J. Vermaak, and A. Blake. Data fusion for visual track-ing with particles.

28 Jul 2009: 2008), doi:10.1016/j.patrec.2008.04.005. and timing each stroke, we were able to estimate the hand labelingtime for one frame to be around 20–25 min. ... 5.35%. 26.41%. 3.34%. 66.52%. 33.07%. 47.97%. 86.90%. 81.01%. 95.90%. 0% 10% 20% 30% 40% 50% 60% 70

28 Jul 2009: and subsequently combinetheir output, by either switching between them [2] or byprobabilistically merging them [8, 15, 18, 20]. ... 20] P. Pérez, J. Vermaak, and A. Blake. Data fusion for visual track-ing with particles.

28 Jul 2009: and subsequently combinetheir output, by either switching between them [2] or byprobabilistically merging them [8, 15, 18, 20]. ... 20] P. Pérez, J. Vermaak, and A. Blake. Data fusion for visual track-ing with particles.