Search
Search Funnelback University
101 -
150 of
300
search results for KaKaoTalk:po03 op |u:mi.eng.cam.ac.uk
where 0
match all words and 300
match some words.
Results that match 1 of 2 words
-
EFFICIENT DECODING WITH GENERATIVE SCORE-SPACESUSING THE EXPECTATION…
mi.eng.cam.ac.uk/~mjfg/Kernel/van_dalen-2013-efficient_decoding.pdf27 Mar 2013: op-timal segmentation into words, the features are found in amortisedconstant time. -
A Simple Technique for Self-Calibration
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/1999-CVPR-self-calibration.pdf13 Mar 2018: This goal is achieved by solving an op-timization problem by numerical techniques, searching di-rectly for the intrinsic parameters of the cameras, instead ofthe indirect search performed by the algorithms -
��������� ��� ������� ��������� ��������� � ���!��� "��…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/jones_icslp94.pdf9 Aug 2005: " $#%&'( $)!,-. /0& / 1324!5!' 678:95;<=%>$? ,A@BC!1. C)DE&@/1(/! C%$1FG GH "! $ "I DJ1(KLDJ@/DJ(G( M14C5(.GNC!1( 4.1(.J&-!C @O!C!1 ( L MPRQFSBT2FU%VC!(/1%W-!C5 GX!Y&G1( ZM@&1)1F:'! 3F Z) HC! L1([F1(A] 1!Y __@: G GGZ$GZ!G?a,@/! : C5 1(/(T-!I@/M ) @I1 -
DISCRIMINATIVE LANGUAGE MODEL ADAPTATION FORMANDARIN BROADCAST SPEECH …
mi.eng.cam.ac.uk/research/projects/AGILE/publications/liu-asru07.pdf26 Mar 2008: Rather than directly op-timize the interpolation weights, their prior distribution and the as-sociated hyper-parameters are optimized. -
Filtering Using a Tree-Based Estimator B. Stenger∗ A. Thayananthan∗…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2003-ICCV-Stenger-filtering.pdf13 Mar 2018: Filtering Using a Tree-Based Estimator. B. Stenger A. Thayananthan P. H. S. Torr† R. Cipolla. University of Cambridge † Microsoft Research Ltd.Department of Engineering 7 JJ Thompson AvenueCambridge, CB2 1PZ, UK Cambridge, CB3 OFB, UK. -
Stereo Coupled Active Contours Tat-Jen Cham Roberto Cipolla…
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/1997-CVPR-Cham-contour.pdf13 Mar 2018: The tworight columns show the update actions whichmust be performed for the desired tracking op-eration. -
Refining Architectures of Deep Convolutional Neural Networks Sukrit…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2016-CVPR-refining-CNN.pdf13 Mar 2018: Please see Fig 1 for an illustration ofthese operations. We do not consider the other plausible op-erations for architectural refinement of CNN; for instance,arbitrary connection patterns between two layers ... 2. We introduce a strategy that starts with -
���������� ��� ����� ��������� ���������� ������� � �…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/logan_euro97.pdf9 Aug 2005: "#$ %$'&() ((,$-(. /0120$34&5 7689 0(. (,&:. ;7!-&:<(, ,< ='(, > (, " >? '@8AB@DCEGFIHIJ HIJLKM,@DNO@QPREISTUJWVXEIJ. Y:Z[]L_badcfehgjiL_lkIeXGmn_porqsBiLcf_piLee+t_liLcuRevwZto[eXidoXxyY:Zz[]{_Ua|cfeIx{gjiL_loteza} _li{cIad[. :IBBXBlGU8/rt5lX5/t -
��������� �� �������������� ����� ���� ����� ����! "#��…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/waterhouse_nips94.pdf9 Aug 2005: E? P@$#WT $ F @E> "!I $&% &> & @ #" OP@Q;R!# @ AW$FH "KR $ W I E$%#@E! ... diiF:«0JN?CF=eH. Ò º D?E"&% & & % %EA D#OP &>? & G%;FH"JLK'& &&O )"E @?& I @XO $!#& @E>.$!#&&?VPPU! " $%&! " -
Utilisation de la cohérence globale entre silhouettes pour…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2006-ARIF-Hernandez.pdf13 Mar 2018: intervals de profondeur vides.Dans le cas de deux vues, les silhouettes correspondantesne seront pas cohérentes s’il existe au moins un rayon op-tique classé S par une des silhouettes -
PoseNet: A Convolutional Network for Real-Time 6-DOF Camera…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2015-ICCV-relocalisation-arXiv.pdf13 Mar 2018: engineering or graph op-timisation. ... This demonstrates that learning with the op-timum scale factor leads to the convnet uncovering a more accuratepose function. -
��������� �� ��������������������������� ���…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/reinhard_ioa98.pdf9 Aug 2005: p1.06C4FL9.p8rtq56&IJ.DwVZL9476GL9/M15UcQy6&OJRÉL93.r0q6&IJ!69.0X69.pK&.01ZL&.>8PSZN1YX[O7/21ZL9K EN OP@: Ì =uQyOJ6.>47rC3YÏ /21L93.K94DRYXo2. L&35.IF4Do2q.K È K TRQ ST : ... OP.t|$L&6C47rtL47oMo¢Or0r0q696&/21UW8/MX53O71.>Ktwu/ML&3L&3.cS[.pKGL!6&.>rt -
Hole Filling Through Photomontage Marta Wilczkowiak∗, Gabriel J.…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2005-BMVC-Wilczkowiak-hole-filling.pdf13 Mar 2018: Now the problem of finding an op-timal replacement of pixels in patch p′ from those in patch m′ is equivalent to finding apath in the graph such that the error -
Department of Engineering 1 Generative Kernels and Score-Spaces…
mi.eng.cam.ac.uk/~mjfg/Kernel/rcv25_2013_y2.pdf9 Sep 2013: σ ,. 〈logl,. αTλll. 〉. (16). Notice that the last element of the tuple now is a scalar value, which makes the op-erations and (which are similar to (15) and straightforward to -
��������� ��� ����������������������������!…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/reinhard_icassp99.pdf9 Aug 2005: ð|Î Óp«ñBòóñÀ«Ñrôò Ï Ñ ÐÌõ ÑÒÃòög ¤µt£ªbM9_á£bª£¥¤>9¤> 9' ø9H£ª9 eÀn >« ¤Q¤+p9bb9b>ª£¡º!ù?úlåÕ1¡ºb br£«H¤e >}9_ábbª£¤>&b£9'»Q£ª¤> a ' 9 »f_| -
Noname manuscript No.(will be inserted by the editor) Using ...
mi.eng.cam.ac.uk/~cipolla/publications/article/2014-IJCV-dense-AAM.pdf13 Mar 2018: Dense op-. tical flow is used to compute pairwise registration and. -
PoseNet: A Convolutional Network for Real-Time 6-DOF Camera…
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/2015-ICCV-relocalisation.pdf13 Mar 2018: engineering or graph op-timisation. ... This demonstrates that learning with the op-timum scale factor leads to the convnet uncovering a more accuratepose function. -
bmvc-99.dvi
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/1999-BMVC-Chesi.pdf13 Mar 2018: 2Gênu£¢£nZ»i{ki£ªn¢a£ªs£iiÓ]i]n¢«sx¢g¤K1 ¡¥£B¢ai«Åig¢]s£iªs¢£g Ói isÅ1{Z£i3ìs«¥Â£pi£¥NM%¤n]kg]iPO ¢ ¡¥s£<-¿! $. iÓ]i]n¢n¥«¥ÅspJKs P I6¥««3¤Kªs£g Óª££g MV OP I V -
����������� ��� ������ ��� ���� �� ������ ������� ���� ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/malis_bmvc99.pdf9 Aug 2005: 2Gênu£¢£nZ»i{ki£ªn¢a£ªs£iiÓ]i]n¢«sx¢g¤K1 ¡¥£B¢ai«Åig¢]s£iªs¢£g Ói isÅ1{Z£i3ìs«¥Â£pi£¥NM%¤n]kg]iPO ¢ ¡¥s£<-¿! $. iÓ]i]n¢n¥«¥ÅspJKs P I6¥««3¤Kªs£g Óª££g MV OP I V -
��������� ������� ������������������������ ���!…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/lawn_tr160.pdf9 Aug 2005: tDvxmKyz{ag/koamn|@p s p}l}op z{ygKa. mm m. dKm pjym gp}@z{pg s kl z{ygKa. ... 4"<4L86"< )&A$"%( OP SOH! "% $ &%& 6 2 /, 1 ,"%5="% w4"< 2 "< 9 4w" 2 wZJ> 4w"%H7L. -
MVA '98 IAPR Workshop on Machine Vision Applications, Nov. ...
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/1998-MVA-invited.pdf13 Mar 2018: This op- timisation takes into account image gradient in- formation and modifies image point coordinates in order t o position primitive edges along image edges. -
������������� ��� ���������������������� � ���� "!$# #�% &…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/gales_tr135.pdf9 Aug 2005: 6 4. 0. 0! -!3# 0. (.a6 bPQ5'C 1'a $7") O-,/'">/, & 4.| (S4S-.,/7 O2a --'">65, - u3 p) 4 4 ':. p,/ 6 C ()(3 (/7?-,/s ()626 OP -' (PF6 -
� ���� � � � � � � � ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/leggetter_icslp94.pdf9 Aug 2005: No. Adaptation Utterances. Speaker IndependentSpeaker DependentSpeaker Adapted. 3 Iu%%&'&6"%q94$h%"Q%{Q%hX&'&%&wJ68(S)9"i(ON"&RJ6'w&>xv%'Op ... f"1(Oo$NfhhfJc6>oNQOI&'0LoNM%hon&('6e1N(OP"&'4wQ%%"Q%)/0'65h%/&N"(ORh65'f"nr(IwQ%h4&('%"y(6y('In]o$hXhfwL(Orzo) -
��������� ��� �������������� ����������� ���…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/drummond_cvpr99.pdf9 Aug 2005: KL ÚM.L þ ÿN OP ÚG QQQQ NR þ£ê!SUTS ÿ þ>ú ÿ. ... K L V LW OP þ ûDÿ˪ÌÍØIÓuÖÒÝÍPÏ V L ÓwÒÍÒÍѤÍÒÒÍÚ!ÖÐ$ÓOÏ Þ ÍÛÍ&ÒÓuÖÐwÒÏÐÑ. -
Statistical Machine Translationand Automatic Speech Recognitionunder…
mi.eng.cam.ac.uk/~wjb31/ppubs/LMathiasDissDec07.pdf16 Feb 2008: with reliable verbatim transcripts. However, accurately transcribed training data. are not always available and manually generating them is not always a feasible op-. ... The uncertainty in speech translation is due to the ambiguity in selecting the op-. -
Multi-Sensory Face Biometric Fusion (for Personal Identification)…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2006-OTCBVS-Arandjelovic-fusion.pdf13 Mar 2018: The optimal. values were found to be2.3 and6.2 for visual data; the op-timal filter for thermal data was found to be alow-passfilterwith W2 = 2.8 (i.e. -
Silhouette-based Object Phenotype Recognition using 3D Shape Priors…
mi.eng.cam.ac.uk/~cipolla/publications/inproceedings/2011-ICCV-Chen-priors.pdf13 Mar 2018: However, the back-projectionfrom 2D to 3D is usually multi-modal, and this results ina non-convex objective function with multiple local op-tima, which is usually difficult to solve. -
KIM et al.: GROWING A TREE FROM DECISION REGIONS ...
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/2010-BMVC-supertree.pdf13 Mar 2018: Huffman coding [17] is related to our op-timisation. It minimises the weighted (by region prior in our problem) path length of code(region). -
A Unifying Resolution-Independent Formulation for Early Vision∗ Fabio …
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/2012-CVPR-Viola.pdf13 Mar 2018: Our implementation has been optimizedfor speed at an algorithmic level by use of second order op-timizers, and by limiting the number of polygon clippingsrequired, but has not been micro-optimized. -
Int J Comput Vis (2012) 100:203–215DOI 10.1007/s11263-011-0461-z…
mi.eng.cam.ac.uk/~cipolla/archive/Publications/article/2012-IJCV-Shallow-trees.pdf13 Mar 2018: path-length. to train data i.e. good generalisation, however, it is not op-timal in classification time. ... 2008). Random kitchen sinks: replacing op-timization with randomization in learning. In Proc. -
��������� ���� �� ������� �� ������ ���� �� ���������� ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/carr_WSCG98.pdf9 Aug 2005: "# $%#&'( )%. ,.-0/21430/-658795:/;4;<=?>"-?@A;CBDFE79G BB2<H=0I!JLKM30/N;4@POQ7?R(;M/S.B;<"TVUWBXYJZ-[70]/N_1C,N-?a< ]8BbA17(,Nc de-SNJZ-BB;4J-0S=gf8-JZXNB;4hCJ1ji,Nc5:/Nkl;4JL@ASNBN=5:/Nkl0;CJL@AS.B=gfU7?d(kP/NJZ_nm0opKKKHqB-S07 K/k7r/NK7 st. a -
��������� �� �� ��� ������������������� ������ ������ "!#�����$� …
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/leggetter_tr181.pdf9 Aug 2005: MMJK?1A<>=@4L =@<KA,;,/81GCR 81MJK2JK80?JKMlAM3,/; ). B. &. &. & & &. 32 OP:IJ5H>OPH A ;=?5A E 6$7I8%:;=8 ;=? @BA. 32 @ OP: <;V8 ;=? @ A. ,W ,/M81A4/,f=@?=. 0,/G ,/?S2 BW;= 2 =@H -
PoseNet: A Convolutional Network for Real-Time 6-DOF Camera…
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/2015-ICCV-relocalisation-arXiv.pdf13 Mar 2018: engineering or graph op-timisation. ... This demonstrates that learning with the op-timum scale factor leads to the convnet uncovering a more accuratepose function. -
Variable-length category-basedn-grams for language modelling T.R.…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/niesler_tr215.pdf9 Aug 2005: Variable-length category-basedn-grams for language modelling. T.R. Niesler and P.C. Woodland. CUED/F-INFENG/TR.215. 30 April 1995. Cambridge University Engineering Department. Trumpington Street, Cambridge, CB2 1PZ. trn@eng.cam.ac.uk. -
��� ������� �� � �������������������� ���� � � �"!�� ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/aiyer_tr55.pdf9 Aug 2005: "! $#&%(')),'. -./10, 2+3, 4 5 67 98:,2+ ,5 ;<2$=?>56/$?"6@! </$>, ,26A B656-. /C0,5 2 D-.FEG'IHKJ. 8:,2+. L 40M'INCOQPR TSU57 <2V/W V! ,X! ,0,/B ;656Y6ACZ8[8K8 @ < V]6 A ,A+_!,:.6RaA+X. b:cedIfg. -
paper.dvi
mi.eng.cam.ac.uk/~mjfg/wang_icassp13.pdf13 Jun 2013: Section 3 discusses op-tions to adapt TANDEM systems. Experiment and results arediscussed in section 4 with the conclusions in section 5. -
CAMBRIDGE UNIVERSITYENGINEERING DEPARTMENT ��������� ��…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/freitas_tr313.pdf9 Aug 2005: jRAKJP7 uq?et 9&;m_&5;q]5 oP;et4optqBN917R?e<9msPBe9dtX%?=<9dl EAKyHg4tO;=>? ... iMAM?e>Ht4JPBmAM?7:>o 9;=9Jq?o 9;=>@t:7<B? Jus<bH A jAOJP7uq?=t 9;0 B AKo<oP;etqAO[I C4?e<9T? -
Joint Uncertainty Decoding for Noise Robust Speech Recognition H. ...
mi.eng.cam.ac.uk/~mjfg/liao_INTER05.pdf19 Dec 2006: Op-erations Room noise from the NOISEX-92 database was addedat the waveform level. -
IB-descriptors.dvi
mi.eng.cam.ac.uk/~cipolla/lectures/PartIB/old/2012-IB-handout3.pdf11 May 2012: The way to achieve this robustness is to utilize in-terest points in computing the descriptor as op-posed to the raw image data. -
SSVM_LVCSR.dvi
mi.eng.cam.ac.uk/~mjfg/sxz_TASLP12.pdf19 Oct 2012: ALGORITHM WITHOUT /WITH OP TIMIS ING θ AND 1-S LACK ALGORITHMSWITHOUT/WITH GAUS S IAN P RIOR(A LG. -
Learning a Kinematic Priorfor Tr ee-Based Filterin g A. ...
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/2003-BMVC-Thayananthan-priors.pdf13 Mar 2018: PQ. RRSST>TU>U VWX>XY>YZ>Z[] >_ >abc. d>defg hi jk. l>lm. n>no>op>pq. rs ttuu. -
��������� �� � ���������������������!…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/ellis_tr122.pdf9 Aug 2005: 1(LQ34?OP(0EL>4G7:=Gs71b<7 3F >!=7}LK0H7 P V g+NMI. ... Gs7bO7I3 Pq13457;<)6dI1L13C?OP().d ).k>47PD6>CL!1S>-bcPÁ>Cm7A347o-L>v)>7 |m7I3471347;<7 34?¤7! -
��������� �� ���� ���������������������������� �����"!#���…
mi.eng.cam.ac.uk/~cipolla/archive/Publications/inproceedings/1994-BMVC-Lawn.pdf13 Mar 2018: "!# $%& %'()#,&-##)./10324 56&-786&9: ;#<: =;#: >?@-<;A66"B9: ;#<: =;#: >? C<DFEGHIJBGKMLONQPLSRLTU(VQLP LXWZYQ[&R]ARR[&T_T_UVALO]QRLOBaNFLPSL=NQU ]bWBcdP ]QNA[eTAdfUf[&ghNQPBiLRU P]QSU]QP LjWBTQYhLdghU[&ThPLBiLP kXWlc&dP -
������� ��� �� ������ ����� ����������������� � ��������� !#"$�…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/hain_icslp98.pdf9 Aug 2005: #"$ %&' )( ,. -/.1024365 78.:9;.=<?>@>@ACBD245EA. FHGJI'KLNMPORQJSUT=VMDWJSXLZY[MD]_aVRQbMcVRSXSXLNMPVRQed=SXfgG4L[[IhS@ViXjk LNlIhfgMPVRQJnmbVopnLNSXSXXjqFHGJI'KLNMPORQJSFHrtsvuxwtyzjRT={|. }@bEEp4[bR;C? ;. -
paper.dvi
mi.eng.cam.ac.uk/~ar527/ragni_is2018a.pdf15 Jun 2018: As described in Section 4 interpolation weights can be op-timised alternatively by maximising the average mapped con-fidence score on the VWB data. -
robust.dvi
mi.eng.cam.ac.uk/~sjy/papers/heyo04.pdf20 Feb 2018: 4.2 Log-Linear Interpolation. Log-linear interpolation has been applied to languagemodel adaptation and has been shown to be equivalentto a constrained minimum Kullback-Leibler distance op-timisation problem(Klakow, -
Multi-domain Neural Network Language Generation forSpoken Dialogue…
mi.eng.cam.ac.uk/~sjy/papers/wgmr16.pdf20 Feb 2018: By op-timising directly against the desired objective func-tion such as BLEU score (Auli and Gao, 2014) orWord Error Rate (Kuo et al., 2002), the model canexplore its output space -
THE CU-HTK MANDARIN BROADCAST NEWS TRANSCRIPTION SYSTEM R. Sinha, ...
mi.eng.cam.ac.uk/~mjfg/sinha_ICASSP06.pdf22 Nov 2006: The op-erating point selected for this work was to use the 1 hour of Englishdata from the TDT4 Mandarin source, along with 10 hours of datafrom the TDT4 English data. -
Combining I-vector Representation and Structured Neural Networks for…
mi.eng.cam.ac.uk/~mjfg/icassp16_wu.pdf5 Apr 2016: Op-tionally, common layers can be introduced before the basis split-ting or after their combination. -
SPEECH RECOGNITION SYSTEM COMBINATION FOR MACHINE TRANSLATION M.J.F.…
mi.eng.cam.ac.uk/~mjfg/gales_ICASSP07.pdf22 Jun 2007: Thus the C2W segmentations differed. Interms of hypotheses combination there are two possible levels to op-erate at.
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.