Article Structure
Abstract
Correct stress placement is important in text-to-speech systems, in terms of both the overall accuracy and the naturalness of pronunciation.
Introduction
In many languages, certain syllables in words are phonetically more prominent in terms of duration, pitch, and loudness.
Background and Task Definition
There is a long history of research into the principles governing lexical stress placement.
Automatic Stress Prediction
Our stress assignment system maps a word, w, to a stressed-form of the word, v‘v.
Stress Prediction Experiments
In this section, we evaluate our ranking approach to stress prediction by assigning stress to spoken and written words in three languages: English, German, and Dutch.
Lexical stress and L2P conversion
In this section, we evaluate various methods of combining stress prediction with phoneme generation.
Conclusion
We have presented a discriminative ranking approach to lexical stress prediction, which clearly outperforms previously developed systems.
Topics
SVM
- We represent words as sequences of substrings, and use the substrings as features in a Support Vector Machine ( SVM ) ranker, which is trained to rank possible stress patterns.Page 1, “Abstract”
- We divide each word into a sequence of substrings, and use these substrings as features for a Support Vector Machine ( SVM ) ranker.Page 1, “Introduction”
- The task of the SVM is to rank the true stress pattern above the small number of acceptable alternatives.Page 1, “Introduction”
- The SVM ranker achieves exceptional 96.2% word accuracy on the challenging task of predicting the full stress pattern in English.Page 1, “Introduction”
- We use a support vector machine ( SVM ) to rank the possible patterns for each sequence (Section 3.2).Page 2, “Automatic Stress Prediction”
- These units are used to define the features and outputs used by the SVM ranker.Page 3, “Automatic Stress Prediction”
- The SVM can thus generalize from observed words to similarly-spelled, unseen examples.Page 3, “Automatic Stress Prediction”
- 3.2 Stress Prediction with SVM RankingPage 3, “Automatic Stress Prediction”
- We adopt a Support Vector Machine ( SVM ) solution to these ranking constraints as described by J oachims (2002).Page 4, “Automatic Stress Prediction”
- We use an SVM because it has been successful in similar settings (learning with thousands of sparse features) for both ranking and classification tasks, and because an efficient implementation is available (J oachims, 1999).Page 4, “Automatic Stress Prediction”
- For our example, pronounce —> ron-no-un-ce, if the SVM chooses the stress pattern, 0-1-0-0, we produce the correct stress-marked word, pronounce.Page 5, “Automatic Stress Prediction”
See all papers in Proc. ACL 2009 that mention SVM.
See all papers in Proc. ACL that mention SVM.
Back to top.
Support Vector
- We represent words as sequences of substrings, and use the substrings as features in a Support Vector Machine (SVM) ranker, which is trained to rank possible stress patterns.Page 1, “Abstract”
- We divide each word into a sequence of substrings, and use these substrings as features for a Support Vector Machine (SVM) ranker.Page 1, “Introduction”
- We use a support vector machine (SVM) to rank the possible patterns for each sequence (Section 3.2).Page 2, “Automatic Stress Prediction”
- Table l: The steps in our stress prediction system (with orthographic and phonetic prediction examples): (1) word splitting, (2) support vector ranking of stress patterns, and (3) pattem-to-vowelPage 3, “Automatic Stress Prediction”
- We adopt a Support Vector Machine (SVM) solution to these ranking constraints as described by J oachims (2002).Page 4, “Automatic Stress Prediction”
See all papers in Proc. ACL 2009 that mention Support Vector.
See all papers in Proc. ACL that mention Support Vector.
Back to top.
gold-standard
- 2) ORACLESYL splits the input word into syllables according to the CELEX gold-standard , before applying SVM ranking.Page 5, “Stress Prediction Experiments”
- The output pattern is evaluated directly against the gold-standard , without pattem-to-vowel mapping.Page 5, “Stress Prediction Experiments”
- 5) ORACLESTRESS: The same input/output as LETTERSTRESS, except it uses the gold-standard stress on letters (Section 4.1).Page 7, “Lexical stress and L2P conversion”
See all papers in Proc. ACL 2009 that mention gold-standard.
See all papers in Proc. ACL that mention gold-standard.
Back to top.
