Search
Search Funnelback University
- Refined by:
- Date: 2019
1 -
10 of
26
search results for TALK:PC53 20 |u:mi.eng.cam.ac.uk
where 0
match all words and 26
match some words.
Results that match 1 of 2 words
-
Automatic Grammatical Error Detection of Non-native Spoken Learner…
mi.eng.cam.ac.uk/~kmk/presentations/ICASSP2019_GED_Knill.pdf30 Sep 2019: 10000. 15000. 20000. 25000. 0%. 5%. 10%. 15%. 20%. 25%. 30%. -
BI-DIRECTIONAL LATTICE RECURRENT NEURAL NETWORKSFOR CONFIDENCE…
mi.eng.cam.ac.uk/~ar527/ragni_icassp2019.pdf5 Feb 2019: Numerous hand-crafted fea-tures have been proposed [20, 21, 22, 23]. In the simplest case, du-ration and word posterior probability can be used as input features.More complex features ... 20] T. Schaaf and T. Kemp, “Confidence measures for -
CONFIDENCE ESTIMATION AND DELETION PREDICTION USINGBIDIRECTIONAL…
mi.eng.cam.ac.uk/~mjfg/ALTA/publications/SLT2018_ragni.pdf31 Aug 2019: 10. 20. 30. 40. 0.9. 0.8. 0.7. 0.6. Data (%). WE. ... thresholding scheme generalised better to the wide-band data. 0 20 40 60 80 1000. -
Towards Learning Orientated Assessment for Non-native Learner Spoken…
mi.eng.cam.ac.uk/~kmk/presentations/ALTA_Sheffield_20190306.pdf8 Mar 2019: 400 hour BULATS training set. 20. AM LM % WER. -
SEQUENCE TEACHER-STUDENT TRAINING OF ACOUSTIC MODELS FOR…
mi.eng.cam.ac.uk/~mjfg/ALTA/publications/wang_slt18.pdf25 Feb 2019: se-quence training can often yield significant performance gains [20].Thus, sequence-level criteria have been introduced into the TS train-ing framework. ... System Target %WERPolish Arabic Viet. French Thai Dutch OverallTS En-graph 20.8 31.4 32.2 22.4 30 -
Applying Deep Learning in Non-native Spoken English Assessment
mi.eng.cam.ac.uk/~kmk/presentations/APSIPA2019_Knill_Keynote.pdf21 Nov 2019: 20/45. Assessment: Gaussian Process [14, 16]. • Gaussian process• non-parametric model based on joint-Gaussian assumption. • ... 16-20, 2017, 2017, pp. -
LEARNING BETWEEN DIFFERENT TEACHER AND STUDENT MODELS IN ASR ...
mi.eng.cam.ac.uk/~mjfg/ALTA/ASRU2019_TS.pdf20 Dec 2019: Furthermore, sequence-level training often yields abetter performance than frame-level training when training towardthe reference transcriptions [20]. ... The diagonal-covariance GMM for AMI-IHM had 20 mixture components perstate and used 13-dimensional -
BI-DIRECTIONAL LATTICE RECURRENT NEURAL NETWORKSFOR CONFIDENCE…
mi.eng.cam.ac.uk/~mjfg/ALTA/publications/ICASSP2019_li.pdf31 Aug 2019: Numerous hand-crafted fea-tures have been proposed [20, 21, 22, 23]. In the simplest case, du-ration and word posterior probability can be used as input features.More complex features ... 20] T. Schaaf and T. Kemp, “Confidence measures for -
To appear Proc. ICASSP. c©2019 IEEE. Personal use of ...
mi.eng.cam.ac.uk/~mjfg/ALTA/publications/Knill_ICASSP2019_AcceptedPaper.pdf3 Mar 2019: Section E is made up of 5x 20 second responses tosub-questions related to an overall topic e.g. ... 20, pp. 37–46, 1960. [15] Mary L. McHugh, “Interrater reliability: the kappa statistic,”Biochemia Medica, vol. -
IMPROVED AUTO-MARKING CONFIDENCE FOR SPOKEN LANGUAGE ASSESSMENT M.…
mi.eng.cam.ac.uk/~mjfg/ALTA/publications/vecchio_slt18.pdf25 Feb 2019: p(x|z,θ) = N(x|fµ(z|θ),fΣ(z|θ)), (20). where fµ(z|θ) and fΣ(z|θ) are the outputs of the DDN,which is parametrised by ... 5769–5779. [20] Lucy Chambers and Kate Ingham, “The BULATS On-line Speaking Test,” Research Notes, vol.
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.