Báo cáo khoa học: "Grounded Language Modeling for Automatic Speech Recognition of Sports Video" doc

... Proceedings of ACL-08: HLT, pages 121–129, Columbus, Ohio, USA, June 2008. c 2008 Association for Computational Linguistics Grounded Language Modeling for Automatic Speech Recognition of Sports ... are learned from large cor- pora of unlabeled video, and are applied to the task of automatic speech recognition of sports video. Results show that grounded la...

Ngày tải lên: 17/03/2014, 02:20

... topic-adapted model for re-ranking within each set of predictions, but found that the forced ordering of the ngram components was overly detrimental to keystroke savings. Backoff models for topic modeling ... either before or after the linear interpola- tion. If the backoff is performed after interpolation, we must also choose whether smoothing (a prereq- uisite for backoff) is pe...

Ngày tải lên: 31/03/2014, 00:20

... performance of statistical translation systems. To solve this problem, this paper proposes a novel method for phrase-based SMT by using a pivot language. To perform translation between languages ... In general, for most of the languages, there exist bilingual corpora between these languages and English since English is an internationally used language. Table 1 shows the inf...

Ngày tải lên: 08/03/2014, 02:21

... frequency of n-gram x in S train qc(x) ← quantisation of c(x) (Eq. 1) for j = 1 to qc(x) do for i = 1 to k do h i (x) ← hash of event {x, j} under h i BF[h i (x)] ← 1 end for end for end for return ... qc(x) ∈ S train for j = 1 to M AXQCOUNT do for i = 1 to k do h i (x) ← hash of event {x, j} under h i if BF[h i (x)] = 0 then return j − 1 end if end for end for The p...

Ngày tải lên: 17/03/2014, 04:20

... see Collins (2004) for more discussion. 3 Linear models for speech recognition We now describe how the formalism and algorithms in section 2 can be applied to language modeling for speech recognition. 3.1 ... possible strings in the language. Markov (n-gram) models are of- ten used for this task, whose parameters are optimized to maximize the likelihood of a large amo...

Ngày tải lên: 23/03/2014, 19:20

... coordinates of an indefinite point P are given in the form: COORD K OF P (REFERRED_TO A)=H where K is a group of possible coordinates, H a set of values for these coordinates and A is the THOUGHT of ... operations of: 1-LOOKING for compatibility of certain features of md,md-l, and md-2. Typical features are superset and part -of relations of md's. A rule may...

Ngày tải lên: 24/03/2014, 05:21

... spo- ken language of Switchboard. Finally, it is clear that the addition of the WSJ data to the NYT data yields improvements only for the specific newswire domain — none of the results change much for ... empirical study of smoothing techniques for language modeling. Technical Report, TR-10-98, Harvard University. Joseph H. Danks and Sam Glucksberg. 1971. Psycho- logical sc...

Ngày tải lên: 30/03/2014, 21:20

... previous research on Ara- bic language modeling and Arabic automatic speech recognition in two other ways. To promote cross- dialectal use of the techniques, we use properties of Arabic morphology that ... decrease in word error rate in Arabic auto- matic speech recognition. 1 Introduction The task of language modeling is to predict the next word in a sequence of wor...

Ngày tải lên: 31/03/2014, 00:20

... tractable amount of time, since according to the Markov as- 2 Often these are more complicated than picking informative features as proposed in this paper. One example of the kind of operator used ... selection is often intuitive, recent work has ex- perimented with automatic feature selection using LDA (Druck et al., 2008). For some of the exper- iments we also use two sets of...

Ngày tải lên: 20/02/2014, 09:20

... values for the family of metrics AEv(α,N), for adequacy scores in MT evaluation scores from the human evaluation, for summaries of 200 and 400 words, respectively (the values of R 2 for summaries ... restricts the domain of questions to be handled to so-called factoid questions. Automatic evaluation of factoid QA is often straightforward, as the number of correct...

Ngày tải lên: 20/02/2014, 16:20