Search
Search Funnelback University
- Refined by:
- Date: 2005
Did you mean economiespast |u:mi.eng.cam.ac.uk?
1 -
50 of
89
search results for Economics test |u:mi.eng.cam.ac.uk
where 1
match all words and 88
match some words.
Fully-matching results
-
MORTGAGE DEFAULT: CLASSIFICATION TREES ANALYSIS David Feldman* and…
mi.eng.cam.ac.uk/~mjfg/local/4F10/Feldman_Gross.pdf15 Nov 2005: Cross-Validation method, and the Test-Sample method. In the former, the learning. ... replace the more costly Cross-Validation misclassification estimation by the. Test-Sample method. Results that match 1 of 2 words
-
The Cambridge University March 2005 Speaker Diarisation System
mi.eng.cam.ac.uk/reports/full_html/sinha_eurospeech05.html/22 Sep 2005: Using just the target test show for Ewas not effective. ] The DER on the. ... E. uses the test data (unsupervised). Concatenation () and MAP adapted ( ) results are given. -
Recent Improvementsin the CUED CTS SU System M. Tomalin ...
mi.eng.cam.ac.uk/research/projects/EARS/pubs/tomalin_mar05mde.pdf28 Apr 2005: increasing mixtures.• Obtain prob from each SU subtype GMM for each feature vector in test data.• Place GMM probs on arcs of lattice and decode as usual. ... Initial GMM results for dev03f test set:. SYSTEM DEL INS SUBS %Err. -
Two-way Cluster Voting to Improve Speaker Diarisation Performance
mi.eng.cam.ac.uk/reports/full_html/tranter_icassp05.html/31 Mar 2005: NIST,. ''Benchmark Tests : Rich Transcription (RT),''. -
A PLSA-based Language Model for Conversational Telephone Speech David …
mi.eng.cam.ac.uk/reports/svr-ftp/mrva_icslp04.pdf10 Jan 2005: gram. The reduction is greater if PLSA’straining text relates to the test set. ... The test set perplexities were calculated on the re-maining half of eval03 (eval03tst). -
The Development of the Cambridge University RT-04 Diarisation System
mi.eng.cam.ac.uk/reports/full_html/tranter_rt04.html/10 Jan 2005: NIST,. Benchmark Tests : Rich Transcription (RT),. -
ADAPTATION OF PRECISION MATRIX MODELS ON LARGE VOCABULARYCONTINUOUS…
mi.eng.cam.ac.uk/reports/svr-ftp/sim_icassp2005.pdf12 Apr 2005: For MLLR mean adaptation, a. SystemAdapt Test Set WER (%)Config eval03 dev04 dev04f. ... This system wasevaluated on botheval03 and dev04 test sets and comparedwith the SATDIAGC system. -
nips7.dvi
mi.eng.cam.ac.uk/reports/svr-ftp/williams_nips2004.pdf9 Aug 2005: This is restrictive in practicein that test data may contain distortions that take it outside the strict ambit of the trainingpositives. ... This can be computed in terms of likelihoods. (1)so then the test becomes (2)where. -
The Cambridge University March 2005 Speaker Diarisation System R. ...
mi.eng.cam.ac.uk/reports/svr-ftp/sinha_eurospeech05.pdf23 Jun 2005: data when processing all the test data togetheror 8.6% when processing the test data shows independently. ... The previous experiment was repeated adding the test dataitself, E, instead of the dev data. -
TWO-WAY CLUSTER VOTING TO IMPROVE SPEAKER DIARISATION PERFORMANCE S.…
mi.eng.cam.ac.uk/reports/svr-ftp/tranter_icassp05.pdf25 Mar 2005: 6. REFERENCES. [1] NIST, “Benchmark Tests : Rich Transcription (RT),” http://www.nist.gov/speech/tests/rt/. ... 2] NIST, “The Rich Transcription Spring2003 (RT-03S) Evaluation Plan, version 4,”http://www.nist.gov/speech/tests/rt/rt2003/spring/docs -
The Cambridge University March 2005 Speaker Diarisation System R. ...
mi.eng.cam.ac.uk/reports/full_html/sinha_eurospeech05.html/paper.pdf22 Sep 2005: data when processing all the test data togetheror 8.6% when processing the test data shows independently. ... The previous experiment was repeated adding the test dataitself, E, instead of the dev data. -
NEW FEATURES IN THE CU-HTK SYSTEM FOR TRANSCRIPTION OFCONVERSATIONAL…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/hain_icassp01.pdf9 Aug 2005: Vocal tract length nor-malisation (VTLN) was applied in both training and test. ... Similar im-provements have been found on other test sets. 7. FULL VARIANCE TRANSFORMS. -
A PRACTICAL PERCEPTUAL FREQUENCY AUTOREGRESSIVE HMMENHANCEMENT SYSTEM …
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/logan_icslp98.pdf9 Aug 2005: They are obtained by optimising the inser-tion penalty and another parameter, the silence probability incre-ment, for each test condition. ... Lynx noisefrom the NOISEX-92 database was added to the test sets at 18dBand 12dB. -
� ������� � ��� ����� ��������� ���������������������� ���…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/niesler_tr258.pdf9 Aug 2005: where4. is retained inE G. when the test succeeds. In practice the value of is fixed (e.g. ... Training- and test-sets were created by splitting thematerial evenly across these topics in the ratio 95:5, resulting in a vocabulary size of 41097 words. -
Recent Developments at Cambridgein Broadcast News Transcription D.Y.…
mi.eng.cam.ac.uk/research/projects/EARS/pubs/kim_rt04.pdf15 Feb 2005: Presentation Overview. • RT03 Broadcast News System Review• Training & Test Data• Improved Acoustic Model Building. – ... Test set # Shows Size Period. dev03 6 3hrs Jan01eval03 6 3hrs Feb01dev04 6 3hrs Jan01dev04f 6 3hrs Nov03. -
Advances in Structural Metadata for RT-04 atCUED M. Tomalin ...
mi.eng.cam.ac.uk/research/projects/EARS/pubs/tomalin_rt04.pdf15 Feb 2005: Training and Test Data for CTS. The following sets of CTS training data were used:. ... Training and Test Data for BN. The following sets of BN training data were used:. -
paper.dvi
mi.eng.cam.ac.uk/reports/svr-ftp/kim_icslp04.pdf10 Jan 2005: The transcription is either known in training orobtained by an initial, non-VTLN, decoding pass for test data.If and are the original and transformed feature vectorsrespectively then the log-likelihood ... The systems were eval-uated on two 3 hour test -
INVESTIGATION OF ACOUSTIC MODELLING TECHNIQUES FOR LVCSR SYSTEMS M.…
mi.eng.cam.ac.uk/reports/svr-ftp/gales_rt04_modelling.pdf19 May 2005: This system wasevaluated on both eval03 and dev04 test sets and comparedwith the SATDIAGC system. ... Adaptation experimentswere conducted based on three 3-hour test sets: eval03, dev04and dev04f. -
SENSORLESS FREEHAND 3DULTRASOUND IN REAL TISSUE: SPECKLE…
mi.eng.cam.ac.uk/reports/svr-ftp/gee_tr510.pdf10 Jan 2005: Subsequently, separate distance estimates wereobtained for each patch in the three test sets. ... To facilitate the noise test (3), axial decorrelation curves werealso calculated for each patch. -
� ������� �� ���������� ���������������� � �� � ������!"� ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/niesler_tr281.pdf9 Aug 2005: and kill-tests respectively, are used to decide upon the strengthof the correlation. ... The kill-test succeeds when 8 X , and the following figure illustrates these conditions. -
Face Recognition with Image Sets Using Manifold Density Divergence ...
mi.eng.cam.ac.uk/reports/svr-ftp/oa214_CVPR_2005_paper1.pdf8 Aug 2005: These meth-ods have achieved very good accuracy on a small number ofcontrolled test sets. ... We therefore useDKL(p(0)||p(i)) as a “distance measure” between trainingand test sets. -
IMPLEMENTATION OF AUTOMATIC CAPITALISATIONGENERATION SYSTEMS FOR…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/kim_icassp02.pdf9 Aug 2005: Particular importance must be given to the effect of words en-countered in the test data which have not been seen in the trainingdata. ... We also used 3 hours of test data from the NIST 1998 Hub-4BN benchmark tests. -
INVESTIGATION OF ACOUSTIC MODELING TECHNIQUES FOR LVCSR SYSTEMS X. ...
mi.eng.cam.ac.uk/reports/svr-ftp/liu_icassp2005.pdf19 May 2005: Among all the adaptivelytrained systems, the SATSPAM outperformed all the other sys-tems on both test sets. ... 3. Three BN test sets were used, each of them contains six 30 minutesbroadcast news shows. -
Recent Improvements in the CUED DiarisationSystem Sue Tranter, Rohit…
mi.eng.cam.ac.uk/research/projects/EARS/pubs/tranter_mar05mde.pdf26 Mar 2005: SID stage - Adding Eval Data to UBM. • New UBMs, BE and BMAPE, were created using the whole test data set.• System gender labels were used. ... no real cheating but against eval rules).• Using just the target show (rather than whole test set) did -
TODO: This is a placeholder. Final title will be filled later.
mi.eng.cam.ac.uk/reports/svr-ftp/liao_interspeech05.pdf26 Sep 2005: Here. P (šn|xt, M̌) P (šn|yt, M̌) (10)the model M̌ is now matched to the test condition rather. ... All results are quotedaverage of three of the four available test sets, Feb’89,89 and Feb’91. -
Progress in English Conversational TelephoneSpeech Transcription Khe…
mi.eng.cam.ac.uk/research/projects/EARS/pubs/sim_sttmar05.pdf12 Apr 2005: Acoustic Model Test Data:– eval03 6 hours (3 hours Switchboard2 Phase 5, 3 hours Fisher)– dev04 3 hours Fisher data. • ... Experimental Setup. • Acoustic model data sets:– Training data: 76 hours h5etrain03sub & 296 hours h5etrain03– Test data -
Appears in Proc. ICASSP-97 c� IEEE 1997 MODELLING WORD-PAIR ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/niesler_icassp97.pdf9 Aug 2005: At each update a t-test isperformed, determining to a desired level of confidence whetherthe measured mean is lower than would be expected were the can-didates to occur independently. ... The following table gives a brief description ofeach corpus, where -
��������� �� ������� ����������� ������� ��� � �� ��! ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/niesler_tr265.pdf9 Aug 2005: The test-set wassubjected to the same preprocessing steps as the training corpus. ... The OOV rate for the describedvocabulary is 0.62% on the test-set. -
PRONUNCIATION MODELING BY SHARING GAUSSIAN DENSITIESACROSS PHONETIC…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/nock_euro99.pdf9 Aug 2005: Steps 6, 3, 4 and 5 are then carried out to estimateand test a pronunciation model. ... in Step 5 (Test) PER WER PER WERNone (Dictionary) 49.1% 49.1% 49.5% 58.9%Tree Pron. -
Contour-Based Learning for Object Detection Jamie ShottonDepartment…
mi.eng.cam.ac.uk/reports/svr-ftp/shotton_iccv05.pdf8 Aug 2005: Test-ing was performed on 164 images containing 193 cars,and164 background images. ... Our technique relies on edge features andhigher-resolution training and test images would certainlyimprove results. -
A NEURAL NETWORK BASED, SPEAKER INDEPENDENT, LARGE…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/wernicke_eurospeech93.pdf9 Aug 2005: Sincethe integration of feedforward and recurrent MLPs have al-ready been shown to be quite successful, the main goal of thisEsprit project is to develop and test these approaches on largescale ... 2. More specifically our current focus is to compare the -
paper-mdeval-v19_revised.dvi
mi.eng.cam.ac.uk/reports/svr-ftp/tomalin_rt04.pdf12 Jan 2005: 10. REFERENCES. [1] NIST, “Benchmark Tests : Rich Transcription (RT),” http://www.nist.gov/speech/tests/rt/. ... ICASSP, March 2005. [19] Linguistic Data Consortium, “Simple Metadata AnnotationSpecification V5,” http://www.nist.gov/speech/tests/rt -
THE DEVELOPMENT OF THE CAMBRIDGE UNIVERSITY RT-04 DIARISATION SYSTEM…
mi.eng.cam.ac.uk/reports/full_html/tranter_rt04.html/paper.pdf10 Jan 2005: nist.gov/speech/tests/rt/rt2003/spring/docs/rt03-spring-eval-plan-v4.pdf , 25th February 2003. ... 3] NIST, “Fall 2004 Rich Transcription (RT-04F) EvaluationPlan,” http://www.nist.gov/speech/tests/rt/rt2004/fall/docs/rt04f-eval-plan-v14.pdf , 30th -
ON THE USE OF SUPPORT VECTOR MACHINES FOR PHONETIC ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/clarkson_icassp99.pdf9 Aug 2005: Agiven test example is then classified as belonging to the classwhose boundary maximizes ')( % ,F-. ... all clas-sifiers. 4. TIMIT EXPERIMENTS. To test the performance of SVMs on a more difficult task we usedthe TIMIT database [3]. -
LOOSELY COUPLED HMMS FOR ASR H.J. Nock� S.J. Young ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/nock_icslp00.pdf9 Aug 2005: We use Isolet1-4 (6240 utterances)to train and speaker-disjoint Isolet5 (1560 utterances) to test. ... The McNemar test finds no significant differences be-tween the Vit or CVit schemes and EM/FL Scheme (at significancelevel. " ). -
johnson05improved.dvi
mi.eng.cam.ac.uk/reports/svr-ftp/johnson_bmvc05.pdf9 Aug 2005: Figure 3: Annotation Performance on training and test data. The training subset consisted of 1667. ... images, or 90% of the data, with the test subset consisting of the remaining 10%, or 214 images. -
Article Submitted to Computer Speech and Language Automatic…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/kim_csl04.pdf9 Aug 2005: This test data isnamed TestBNAcoustic98. TestBNAcoustic98 comprises 3 hours of acoustic data andthe transcription. ... Table 8 shows the result of the capitalisation generation system based on NE classes andpunctuation marks for these test conditions. -
THE USE OF RECURRENT NEURAL NETWORKSFOR CLASSIFICATION T. L. ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/burrows_nnsp94.pdf9 Aug 2005: g g g g g g. decisionthreshold. Figure 7: Test patterns and decision boundary limits for a ‘g–d’ classifier: a) w too small,b) w too large. ... The recurrent connectionthus updates our estimate of the priors, depending on the previous context, y(t 1 -
SVMS, SCORE-SPACES AND MAXIMUM MARGIN STATISTICAL MODELS M.J.F. Gales …
mi.eng.cam.ac.uk/~mjfg/BeyondHMM.pdf15 Jun 2005: The held-out dataset for these experiments wasthe eval03 test set, which consists of 6 hours of data. ... To examine performance on held-out data the eval03 test setwas used. -
CLASSIFICATION USING HIERARCHICALMIXTURES OF EXPERTS S.R.Waterhouse…
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/waterhouse_hme.pdf9 Aug 2005: In this paper weextend the HME to classification and results are reported for three commonclassification benchmark tests: Exclusive-Or, N-input Parity and Two Spirals. ... The points in the test set are offset vertically from the points in the learning -
TPAMI0110-0304-1 1..13
mi.eng.cam.ac.uk/reports/svr-ftp/williams_pami2005.pdf20 Apr 2005: A validator which uses a classifier to test the regionof the image described by the current state estimate. ... 75 percent thenumber of cycles required by boosting for these. short test sequences (including training the expert). -
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: As the complexity of the tree increases, the test-set perplexity moves through a global minimum. ... Test-set perplexityUnigram at 36.15 and 36.14 for training- and. test-set respectively. -
A Face Recognition System for Access Control using Video ...
mi.eng.cam.ac.uk/reports/svr-ftp/arandjelovic_PR06.pdf21 Dec 2005: also see [27]). In AFR tests, such methods are usually outperformed by meth-. ... the data as test input. In all tests, both training data for each person in the gallery,. -
STRUCTURAL METADATA RESEARCH IN THE EARS PROGRAM Yang Liu1,5 ...
mi.eng.cam.ac.uk/reports/svr-ftp/tomalin_icassp05.pdf12 May 2005: been introduced, with NIST reporting re-sults with the Wilcoxon signed rank test for speaker-level averagescore differences. ... Woodland, “SU detection forRT-03F at Cambridge University,” http://www.nist.gov/speech/tests/rt/rt2003/fall/presentations/ -
A Probabilistic Framework for Perceptual Grouping of Features for ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/yow_fg96_1.pdf9 Aug 2005: Facial features are marked by hand and the al-gorithm is run through these test images, making the nec-essary measurements to define each class space. ... d) Proba-bility = 0.9468. 8. Results. We test the algorithm on 100 256x256 images taken -
Edge tracking for motion segmentation anddepth ordering P. Smith, ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/smith_bmvc1999.pdf9 Aug 2005: An implementation is outlined and demonstrated on test sequences con-taining two motions. ... The first example considerstwo frames from the “foreman” sequence, commonly used to test motion segmentationschemes. -
RT-04 MDE Evaluation Systems at CUED M. Tomalin, S. ...
mi.eng.cam.ac.uk/research/projects/EARS/pubs/tomalin_overall_rt04.pdf15 Feb 2005: General System Architecture. The SMD systems used same generic architecture:TEST DATA. -
Automatic Face Recognition for Film Character Retrieval in…
mi.eng.cam.ac.uk/reports/svr-ftp/oa214_CVPR_2005_paper2.pdf8 Aug 2005: 0.85. 0.9. 0.95. 1. Test index. Ran. k or. derin. g sc. ... 0.86. 0.88. 0.9. 0.92. 0.94. 0.96. 0.98. 1. Test index. Ran. -
��������� �� �� ������ ���� ����� ������� ������������� ������� ...
mi.eng.cam.ac.uk/reports/svr-ftp/auto-pdf/renals_icsi92-081.pdf9 Aug 2005: Evaluation is performed over various test sets, typically containing 300 sentences from 10–12new speakers. ... The previous bestresult on this test set was 3.6% reported by CMU. -
A single-frame visual gyroscope Georg Klein and Tom…
mi.eng.cam.ac.uk/reports/svr-ftp/klein_drummond2005BMVC.pdf14 Sep 2005: These cases are discussed in Section 5. Figure 2: Rotation center placement results for four test scenes (un-blurred in top row.). ... Figure 2 shows results the algorithm’s choice of rotation center for a number of differ-ent motions in four test
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.