Search
Search Funnelback University
- Refined by:
- Date: 2008
1 -
41 of
41
search results for TALK:PC53 20 |u:mi.eng.cam.ac.uk
where 0
match all words and 41
match some words.
Results that match 1 of 2 words
-
mi.eng.cam.ac.uk/research/projects/VideoNormals/PAMI_data/seq/jacket/l…
mi.eng.cam.ac.uk/research/projects/VideoNormals/PAMI_data/seq/jacket/lights_gamma_2.20.txt30 Sep 2008: 2.2019738478183761e00 -4.2620057830042271e01 -1.5599458966455859e02 6.5258450636461873e01 8.3216350714224383e00 7.5519842935081215e01 -1.1641083085291052e02 7.1224592163225097e01 3.1763254589209186e01 -3.2680490641280562e00 -1.3432772310200104e02 -
mi.eng.cam.ac.uk/research/projects/VideoNormals/PAMI_data/seq/pirate%2…
mi.eng.cam.ac.uk/research/projects/VideoNormals/PAMI_data/seq/pirate%20shirt/lights_gamma_2.20.txt30 Sep 2008: 6.4928927076747556e00 -9.9358755589829819e00 -9.4247989273443963e01 7.4312114703253336e00 -1.7854247997132028e01 -4.4459062470945497e01 -8.5285874906107836e01 2.8719901889227941e00 -9.8648463913923692e00 4.4251076596636814e01 -8.7960816808938503e01 4 -
Model-Based Approaches to Robust SpeechRecognition Mark Gales with…
mi.eng.cam.ac.uk/~mjfg/talk_kcl.pdf13 Jun 2008: Single-Gaussian Approximation. 20 10 0 10 20 30 400. 0.02. 0.04. ... King’s College London Seminar 20. Model-Based Approaches to Robust Speech Recognition. -
Building Multiple Complementary Systems using Directed Decision Trees …
mi.eng.cam.ac.uk/~mjfg/breslin_INTER07.pdf11 Jan 2008: System bnat05 bnat06 bcat06BASELINE S0 20.2 31.4 42.7. DIRECTED D1 20.4 31.0 42.0D2 20.4 30.7 41.9. ... System bnat05 bnat06 bcat06BASELINE S0 20.2 31.4 42.7. DIRECTED D1 20.7 31.4 42.4D2 20.7 31.2 42.1. -
Model-Based Approaches to Robust SpeechRecognition Mark Gales with…
mi.eng.cam.ac.uk/~mjfg/mjfg_edin08.pdf6 Oct 2008: Cambridge UniversityEngineering Department. 20. Model-Based Approaches to Robust Speech Recognition. Adapting SVMs to Noise Conditions. • ... 1. initial noise model from first and last 20 frames2. hypothesis estimated using compensated models with -
Rank-HCI
mi.eng.cam.ac.uk/~cipolla/publications/invitedTalk/2005-Rank-HCI.pdf27 Aug 2008: Detecting frontal faces. 0 10 20 30 40 50 60 70 80. -
Face and hand detection - ICT
mi.eng.cam.ac.uk/~cipolla/publications/invitedTalk/2003-ICT-Hands-Faces.pdf27 Aug 2008: Detecting frontal faces. 0 10 20 30 40 50 60 70 80. -
Derivative and Parametric Kernels for Speaker Verification C.…
mi.eng.cam.ac.uk/~mjfg/longworth_inter07.pdf8 Jan 2008: 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. %). Baseline φ φλ+ φ. -
Rank-HCI
mi.eng.cam.ac.uk/~cipolla/archive/Publications/invitedTalk/2005-Rank-HCI.pdf27 Aug 2008: Detecting frontal faces. 0 10 20 30 40 50 60 70 80. -
Face and hand detection - ICT
mi.eng.cam.ac.uk/~cipolla/archive/Publications/invitedTalk/2003-ICT-Hands-Faces.pdf27 Aug 2008: Detecting frontal faces. 0 10 20 30 40 50 60 70 80. -
DISCRIMINATIVE LANGUAGE MODEL ADAPTATION FORMANDARIN BROADCAST SPEECH …
mi.eng.cam.ac.uk/~mjfg/liu-asru07.pdf11 Jan 2008: CNN 7.24 12.73 20.48 224.2. Table 5. Model size and text source for English component LMs. ... pp lin eql 72.20 75.26 80.35. nbest linpp 72.21 75.25 80.37eql 72.22 75.31 80.52. -
slides.dvi
mi.eng.cam.ac.uk/~mjfg/gales_AUG08.pdf15 Aug 2008: Cambridge UniversityEngineering Department. August 2008 20. Instantaneous and Discriminative Adaptation for Automatic Speech Recognition. -
Discriminative Classifiers with Generative Kernels for Noise Robust…
mi.eng.cam.ac.uk/~mjfg/gales_INTER08.pdf16 Jun 2008: from 20 to 5 dB,combined with the 4 different additive noise sources N1 to N4,subway, babble, car and exhibition hall. ... A set of 20 confusable digit-pairs were selected based onthe overall confusion matrix for the 16 noise conditions. -
- + Code Breaking for Automatic Speech ...
mi.eng.cam.ac.uk/~wjb31/ppubs/VenkataramaniThesisTalk05.pdf16 Feb 2008: Each word is modeled by a left-to-right 20 state HMM, 12 mixtures per state. ... 0.05 35.3 2.82. 0.10 37.9 2.35. 0.20 41.1 2.06. 0.30 43.2 2.00. -
3D computer model emerges from a pile of snaps - info-tech - 23…
mi.eng.cam.ac.uk/~cipolla/archive/Public-Understanding/2005-NewScientist-3Dmodels.pdf10 Sep 2008: 20:57 09 September 2008. Explore by Subject> Health> Fundamentals> Being Human> Living World> Opinion> Sex and Cloning. -
CAMBRIDGE UNIVERSITY ENGINEERING DEPARTMENT DISCRIMINATIVE…
mi.eng.cam.ac.uk/~mjfg/gales_tr605.pdf13 Aug 2008: 5 Results. The performance of the proposed scheme was evaluated on the AURORA 2 task [20]. ... Two forms of front-end were examined, one HTK-based, the other ETSI-based [20]. -
DEVELOPMENT OF A PHONETIC SYSTEM FOR LARGE VOCABULARY ARABICSPEECH ...
mi.eng.cam.ac.uk/~mjfg/gales-asru07.pdf11 Jan 2008: System Seg Test Set WER (%)bnat06 bcat06. P2a-G3 LM3 CU1 21.1 27.6BBN 20.4 27.3. ... P2b-G3 LM4 CU1 21.3 27.8BBN 20.6 27.6CUa P3a-G3 LM3 CU1 20.4 27.2BNa BBN 19.9 26.9CUb P3b-V3 LM4 CU1 20.0 -
CAMBRIDGE UNIVERSITY ENGINEERING DEPARTMENT DISCRIMINATIVE…
mi.eng.cam.ac.uk/~mjfg/mjfg_tech605.pdf6 Nov 2008: 5 Results. The performance of the proposed scheme was evaluated on the AURORA 2 task [20]. ... Two forms of front-end were examined, one HTK-based, the other ETSI-based [20]. -
Covariance Modelling for Noise-Robust Speech Recognition R. C. van ...
mi.eng.cam.ac.uk/~mjfg/vandalen_INTER08.pdf16 Jun 2008: 20 I. 2. –24. yst1yst. yst1. 35 = Dyet (13). where D is the dynamic coefficient matrix and yet is the vectorof static coefficients in the appropriate window. -
K. Yu, M.J.F. Gales and P.C. Woodland Cambridge University ...
mi.eng.cam.ac.uk/~mjfg/yu-interspeech07.pdf11 Jan 2008: data. 0.65 0.7 0.75 0.8 0.85 0.9 0.95 10. 10. 20. 30. -
Comparison of Freehand 3D UltrasoundCalibration Techniques Using a…
mi.eng.cam.ac.uk/reports/svr-ftp/hsu_tr579.pdf5 Sep 2008: From the geometric model of the stylus, xL = rL (0, 0, 20)t. ... Ultrasound in Medicine & Biology, 20(9):923–936, 1994. A. Fenster, D. B. -
This is a placeholder. Final title will be filled later.
mi.eng.cam.ac.uk/~mjfg/yu-asru05.pdf11 Jan 2008: 20). In the VBM step, the auxiliary function for q(T |n) is similar toequation 17 except for using the nth component of p(T ) and q(T ).Note that q( ... The component weight of theGMM, q(n), is updated using equation 20. -
A QUALITY-GUIDED DISPLACEMENTTRACKING ALGORITHM FOR ULTRASONIC…
mi.eng.cam.ac.uk/reports/svr-ftp/chen_tr593.pdf17 Jan 2008: The focus depth was set at 20 mm, the centre of the frame. ... Lesion detection in simulated elastographicand echographic images: a psychophysical study. Ultrasound in Medicine and Biology, 20:877–891, 1994. -
Minimum Bayes-Risk Techniques forAutomatic Speech Recognition and…
mi.eng.cam.ac.uk/~wjb31/ppubs/SKumarPhDefense2004.pdf16 Feb 2008: dijj′ =. . . . 0 POS(ej ) = POS(ej′ ). 1 otherwise. MBR Techniques in Automatic Speech Recognition and Machine Translation – p.20/33. -
Automatic Speech Recognition and Statistical Machine Translation…
mi.eng.cam.ac.uk/~wjb31/ppubs/LMathiasDefenseDec07.pdf16 Feb 2008: 32. 34. 36. 38. 40. 42. 44. 8 10 12 14 16 18 20 22 24. ... Hyper-Parameter Tuning. 30.2 30.4 30.6 30.8. 31 31.2 31.4. 0 2 4 6 8 10 12 14 16 18 20. -
CONSENSUS NETWORK DECODING FOR STATISTICAL MACHINE TRANSLATIONSYSTEM…
mi.eng.cam.ac.uk/research/projects/AGILE/publications/sim-icassp07.pdf26 Mar 2008: System2004 2005. TER BLEU TER BLEU. ISI Phrase 55.63 35.89 56.792 33.85Hiero 53.64 39.39 54.81 37.06ISI Syntax 54.20 39.92 55.33 ... 20 47.76ISI Hiero 40.53 54.54 42.21 46.49ISI Phrase 41.94 52.35 43.09 45.21ISI Syntax 42.96 52.36 45.00 44.11MBR-BLEU -
DISCRIMINATIVE LANGUAGE MODEL ADAPTATION FORMANDARIN BROADCAST SPEECH …
mi.eng.cam.ac.uk/research/projects/AGILE/publications/liu-asru07.pdf26 Mar 2008: CNN 7.24 12.73 20.48 224.2. Table 5. Model size and text source for English component LMs. ... pp lin eql 72.20 75.26 80.35. nbest linpp 72.21 75.25 80.37eql 72.22 75.31 80.52. -
Bitext Alignment for Statistical Machine Translation
mi.eng.cam.ac.uk/~wjb31/ppubs/YDengDefenseDec05.pdf16 Feb 2008: 4 Conclusions. Y. Deng (Johns Hopkins) Bitext Alignment for SMT 20 / 42. ... 20. 21. 22. 23. 24. 25. Coverage. eval02eval03eval04. 1 2 3 4 523.2. -
NONPARAMETRIC SURFACE REGRESSIONFOR STRAIN ESTIMATION J. E. Lindop,…
mi.eng.cam.ac.uk/reports/svr-ftp/lindop_tr598.pdf6 Mar 2008: However, for practical application we need. 5. 20 40 60 80 100sample number. ... a). 20 40 60 80 100sample number. ry=24 a. ry=44 a. -
DEVELOPMENT OF A PHONETIC SYSTEM FOR LARGE VOCABULARY ARABICSPEECH ...
mi.eng.cam.ac.uk/research/projects/AGILE/publications/gales-asru07.pdf26 Mar 2008: System Seg Test Set WER (%)bnat06 bcat06. P2a-G3 LM3 CU1 21.1 27.6BBN 20.4 27.3. ... P2b-G3 LM4 CU1 21.3 27.8BBN 20.6 27.6CUa P3a-G3 LM3 CU1 20.4 27.2BNa BBN 19.9 26.9CUb P3b-V3 LM4 CU1 20.0 -
A HYBRID DISPLACEMENTESTIMATION METHOD FOR ULTRASONIC ELASTICITY…
mi.eng.cam.ac.uk/reports/svr-ftp/chen_tr615.pdf13 Nov 2008: The former ranges fromexternal palpation [22] to internal ultrasonic radiation force [26], while the latter might involveimaging modalities as diverse as MRI [20] and ultrasound [12]. ... IEEE Transactions on Ultrasonics, Ferro-electrics, and Frequency -
tech.dvi
mi.eng.cam.ac.uk/reports/svr-ftp/hsu_tr584.pdf5 Sep 2008: I(A, B) =H(A) H(B). H(A, B), (20). where H(A) and H(B) denote the marginal entropies of the images and H(A, B) represents theirjoint entropy. -
sig-004.dvi
mi.eng.cam.ac.uk/~mjfg/mjfg_NOW.pdf19 Mar 2008: domain the total likelihood is calculated as log p(Y |w) α log(P (w)) β|w| where αis typically in the range 8–20 and β is typically in the ... Then,. P (w) =K. k=1. P (wk|ck)p(ck|ck1,. , ckN 1). (2.20). -
Uncertainty Decoding forNoise Robust Speech Recognition Hank Liao…
mi.eng.cam.ac.uk/~mjfg/thesis_hl251.pdf17 Sep 2008: Thenumber of filters can vary, usually between 20-40, where more filters are used with a largerbandwidth. ... Hence thefilterbank reduces the normal FFT window of 256 points to a set of 20-40 smoothed filterbankcoefficients or channels. -
Augmented Statistical Models for Classifying Sequence Data Martin…
mi.eng.cam.ac.uk/~mjfg/thesis_ml362.pdf16 Jan 2008: 20 tables. ii. Summary. Many important applications, including speech recognition, document categorisation and. ... xt = Aθt xt1 wθt (2.20). ot = Cθt xt vθt. -
Statistical Machine Translationand Automatic Speech Recognitionunder…
mi.eng.cam.ac.uk/~wjb31/ppubs/LMathiasDissDec07.pdf16 Feb 2008: class. 59. 5.4 Chinese-English set: Effect of entropy regularization on translationperformance α = 3, Nc = 20. ... systems [20, 30, 25]. In the AT&T approach, the translation component is decom-. -
thesis.dvi
mi.eng.cam.ac.uk/~wjb31/ppubs/VenkataramaniDiss05.pdf16 Feb 2008: 20. loss for different words is always the same irrespective of the words being compared. -
Generation and Combination ofComplementary Systems for Automatic…
mi.eng.cam.ac.uk/~mjfg/thesis_cb404.pdf9 Jul 2008: Generation and Combination ofComplementary Systems for. Automatic Speech Recognition. Catherine Breslin. Cambridge University Engineering Departmentand. Darwin CollegeJune 23, 2008. Dissertation submitted to the University of Cambridgefor the degree -
Minimum Bayes-Risk Techniquesin Automatic Speech Recognition and…
mi.eng.cam.ac.uk/~wjb31/ppubs/ShankarKumarDiss04.pdf16 Feb 2008: Minimum Bayes-Risk Techniquesin Automatic Speech Recognition. and Statistical Machine Translation. Shankar Kumar. A dissertation submitted to the Johns Hopkins University in conformity with the. requirements for the degree of Doctor of Philosophy. -
Bitext Alignment forStatistical Machine Translation Yonggang Deng A…
mi.eng.cam.ac.uk/~wjb31/ppubs/YDengDissertationDec05.pdf16 Feb 2008: 9, 26, 20], while the most complex, finest problems take place at the word level. -
��������� �� � ������������������� ��� ��� �!…
mi.eng.cam.ac.uk/~wjb31/ppubs/VlasiosDoumpiotisDissOct04.pdf16 Feb 2008: "$#&%'(),.-%/01 " 2-3 2 42 56879;:<=1) >?1 01@ " ABC". DE F(GH42IJKC7 #%L88NM :O88 F1@ ". PQSRUTWVYXUT&ZX[1]_VYXCaVYT.
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.