Search

He has authored and co-authored more than 200 papers.

12 Sep 2022: to rims # similarity registration canonical_mc3_6756_lmarked.ply individual.ply n/a n/a 0 0 0 999 0 200 1 1 -1 0 0 1 0 # similarityTPS registration ... n/a n/a 0 0 0 999 0 200 10 0 -1 0 1 1 0 # map canonical_mc3_6756_lmarked.ply individual.ply n/a

12 Sep 2022: 2 = to rims # register lr_femur.wrl hr_femur.wrl n/a n/a 0 0 0 999 0 200 1 1 -1 0 0 0 0 # review lr_femur.wrl ... It performs 200 iterations in "review" mode, whereby the surfaces are loaded and spun around for the user to inspect.

13 Mar 2018: li’. 0 200 400 600 800 1000 1200 1400 1600 1800 2000!10! ... 0 200 400 600 800 1000 1200 1400 1600 1800 2000!10!

23 Feb 2002: s1. -. -. 12. 26. 59.11. 57.14. 54.04. 50.65. s1. 200. ... 10. -. -. 50.76. 49.42. 52.91. 51.67. s1. 200. 10. 8.

20 Feb 2018: Bearing in mind that the initial parts of. 1The performance deteriorated after 200 dialogues due to an in-crease in speech understanding errors. ... Table 2: Word error rate for different domainsDomain #Adaptation diags #Diags WERSFCore 200 399 15SF1Ext

20 Feb 2018: 0. 50. 100. 150. 200. 1 2 3 4 5 6 7 8 9 10 11 12. ... basis. 0. 50. 100. 150. 200. 1 5 9 13 17 21 25 29 33.

19 Jan 2018: rate thanthe 200 frames-per-second typical in many sys-tems.

12 Sep 2022: 2 = to rims # aspect ratio registration canonical_femur_5580.wrl individual.wrl n/a n/a 1 0 0 999 0 200 7 1 -1 0 0 1 0 # similarityLAD registration ... It performs 200 iterations in "review" mode, whereby the surfaces are loaded and spun around for the

A value of 200 will sample at twice this density, which should always be sufficient.

28 Apr 2005: Error Analysis. c.200 DEL errors analysed by hand:. • 40.7 %: before asyndetic clause boundaryEx: they destroyed all the national monuments () he destroyed a large area. • ... c.200 INS errors analysed by hand:. • 26.2 %: before potential

10 Dec 2003: 0. 0.5. 1. 1.5. 2. 0 200 400 600 800 1000. % ... 0. 0.2. 0.4. 0.6. 0.8. 1. 1.2. 1.4. 1.6. 0 200 400 600 800 1000.

7 Oct 2023: mark edges. 0 200 400 600 800 1000 1200 1400 1600 1800 2000. ... Ker. nel. 0 200 400 600 800 1000 1200 1400 1600 1800 2000.

19 Jul 2006: Gaussianise the data for each speaker:. 20 15 10 5 0 5 10 15 200. ... 0.01. 0.02. 0.03. 0.04. 0.05. 0.06. 0.07. 0.08. 0.09. 0.1. 20 15 10 5 0 5 10 15 200.

14 May 2018: mark edges. 0 200 400 600 800 1000 1200 1400 1600 1800 2000. ... Ker. nel. 0 200 400 600 800 1000 1200 1400 1600 1800 2000.

12 May 2016: A1 A2 B1 B2. 0. 200. 400. 600. 800. 1000. 0 2 4 6 8 10 12 14 16. ... System HL-dim Training Data. % Error. KNN - SUP 20.8 RNNLM 100 17.5 RNNLM 200 Semi-SUP 9.3.

13 Mar 2018: Images 1 and 5200 0 200. 400. 200. 0. 200. Images 2 and 3. ... 200. Images 4 and 5. Figure8: Imagesof the ellipsesandepipolarlines after convergence.Eachfigure showsthe overlappingof images and , for -GP& 555 # and ¤@@P 555.

18 May 2017: mark edges. 0 200 400 600 800 1000 1200 1400 1600 1800 2000. ... Ker. nel. 0 200 400 600 800 1000 1200 1400 1600 1800 2000.

27 Aug 2008: Horse detection. ROC curve - 3.4% EER. 100 training images. 200 test images.

23 Mar 2004: Key Issues: 2004 Data. • New data planned for collection in 2004 includes– more Fisher English CTS data– 200 hours of Levantine Arabic CTS data– 200 hours of Mandarin CTS data– English

21 Feb 2022: Speaking Time Versus Learner Progression. 0. 100. 200. 300. 400. 500.

21 Sep 2006: Furthermore, since the size of the ROIs may vary from one videoto the other, the final MGO images are resized to the corresponding refined images withstandard size (which is 200 200 ... 3.2 Dimension Reduction by PCA. In order to reduce the number of

14 Sep 2005: Here it is set to 640 pixels. 0 200 400 600 800 10000. ... 50. 100. 150. 200. 250. position. inte. nsi. ty. Maximalgradient ramp5.

20 Feb 2018: As training data, the LEGO corpus [28] is used which con-sists of 200 dialogues (4,885 turns) from the Let’s Go bus infor-mation system [29]. ... Each turnof these 200 dialogues has been annotated with IQ (represent-ing the quality of the dialogue up

13 Mar 2018: 000000. 0"/20. 0. 00#$0. =. 0. $#$0000. 00. ( not! )0#"/200.

8 Aug 2005: 400 200 0 200 4000. 0.01. 0.02. 0.03. 0.04. 0.05. 0.06. ... 100 0 100 200 3000. 0.02. 0.04. 0.06. 0.08. x. p(x).

25 May 2004: Gaussianisation. • Transform any distribution to standard Gaussian N(0, I). 20 15 10 5 0 5 10 15 200. ... 0.01. 0.02. 0.03. 0.04. 0.05. 0.06. 0.07. 0.08. 0.09. 0.1. 20 15 10 5 0 5 10 15 200.

20 Feb 2018: 4. -2. 0. 2. 4. 6. 8. 10. 12. 14. 0 200 400 600 800 1000.

20 Feb 2018: 100 200 300 400 500 600 700 8000. 50. 100. 150. ... 200. 250. 300. Time index (ms) of "And then". F0. extr.

12 May 2016: 0 200 400 600 800 1000 12000. 0.1. 0.2. 0.3. 0.4.

12 May 2016: Speaking Time Versus Learner Progression. 0. 100. 200. 300. 400. 500.

Computer Vision and Image Understanding, 148:193-200, July 2016. K Chaiyasarn, TK Kim, F Viola, R Cipolla and K Soga.

10 Aug 1999: 100. 150. 200. 250. 300. 350. 400. a)0 100 200 300 400. ... 0. 50. 100. 150. 200. 250. 300. 350. 400. b). Figure 3: Transfer of points and lines.

8 Aug 2005: Figure 2:Parallax used to cluster input face images (a). The distributions ofη (1) for the three clusters, computedfrom 200 manually labelled frames is shown in (b). ... 20. 40. 60. 80. 100. 120. (a) (b). 0 100 200 300 400 500 600 700 800 900 10000.

20 Feb 2018: 2. 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000.

4 Aug 2014: 200. 0. 200. 400. 600. 800. 1000. 1200. 1400. distance across cortex (mm). ... 100. 150. 200. 250. 300. 350. 400. 450. 500. true mass (mg/cm2).

13 Mar 2018: Relative motion of an apparent contour. 118. a. 200-. 150-. 100 -. so •. yS S 50. y S > S 100 •. R/inm / B. / / / A. position error/mm. bB. A. - ... aMO -. 200. 150 •. 100. -2.5 -2.0. 1.5 -1.0 -OS.

21 Sep 2006: 0 50 100 150 200 250 300 350 4000.25. 0.2. 0.15. ... Fea. ture. Val. ueBatch OSMINC OSM. 50 100 150 200 250 300 350 4000.

2 Mar 2009: word - - 150.3 69.7 209.8. class. 50 166.6 76.1 217.9exchange 100 164.4 74.8 216.6algorithm 200 160.9 73.5 214.8. ... 400 159.7 73.0 213.950 165.0 73.6 213.5. weight 100 163.5 73.4 213.1merge 200 162.4 73.1 213.0.

13 Mar 2018: 50 100 150 200 250 300 350. 50. 100. 150. 200. ... 200. 250. Figure 3. Examples of the covariances in image 2, for inliersto two motion hypotheses.

20 Feb 2018: This yielded about13K utterances, one for each DA, which is much more difficult than the previous two domains (5.1Kutterances, 200 distinct DAs).

20 Feb 2018: The results are given in Fig. 2. 0 200 400 600 800 1000 1200 1400 1600Training dialogues. ... based adaptation can restore a system tooptimal performance within 200 to 300 adaptationdialogues.

30 Mar 2007: m). 0 50 100 150 200 2503. 2.5. 2. 1.5. 1. ... ated. tilt. angl. e (d. egre. es). 0 50 100 150 200 2501.5.

23 Sep 2003: The Hub-5 systems developedin the past few years ran between 200 and 300xRT and em-ployed 6 to 8 rescoring branches for system combination. ... pass run timeCoding Segmentation 0.068 xRTPass1 0.890 xRTVTLN 0.387 xRTPass2 3.178 xRTPass3.1 Adapt 1.422

20 Feb 2018: The results indicate that the SL-model couldwork quite well with humans, but was improved byRL on the 200 training dialogues.

20 Feb 2018: 33.5. 34. 34.5. 35. 35.5. 36. 36.5. 37. 37.5. 0 200 400 600 800 1000 1200 1400 1600 1800 2000.

5 May 2000: triphone quinphoneHub4 eval’97 0.302 0.163Hub5 dev’98 0.191 -0.026Hub5 eval’98 0.104 -0.200. ... triphone quinphonepost tree post tree. time dep. 0.104 0.234 -0.200 0.188confnet 0.000 0.213 -0.396 0.198.

20 Feb 2018: NonNativeA: Chinese sentences 11-20 and 31-40 spoken bythe first 10 native English speakers, 200 utterances in to-tal.

26 Apr 2022: mark edges. 0 200 400 600 800 1000 1200 1400 1600 1800 2000. ... Ker. nel. 0 200 400 600 800 1000 1200 1400 1600 1800 2000.

9 Aug 2005: Successes. Collisions. Figure 4: Performance improvement with training. In thefirst 200 cycles, only the three elemental tasks were pre-sented to the agent. ... During the first 200 cycles, the Q and gating modules wereupdated in every time-step.