Prominence prediction

Prominence-prediction algorithms follow the same basic approaches as phrase-break prediction, where we have simple deterministic algorithms, sophisticated deterministic algorithms and data-driven algorithms. 

More recently a number of approaches have been proposed which combine the advantages of decision tree approaches (use of heterogeneous features, robustness to curse of dimensionality) and the HMM approach (statistical, global optimal search of sequences). In addition, there has been somewhat of a re-awakening of use of syntactic features due to the provision of more robust parsers. Rather than attempt an explicit model of prosodic phrasing based on trying to map from the syntax tree, most of these approaches use the syntax information as additional features in a classifier [508], [209], [257]. 

---

Concerning the slow dynamics below the crossover temperature Tc, the predictive power of the theory seems to be rather limited. In particular, the emergence of intrinsic slow secondary processes, which seems to be associated with the dynamic crossover in the experimental spectra, is not contained even in the extended versions of the theory consequently, the slow dynamics spectrum is not reproduced correctly. In this respect, the extended theory introducing the hopping mechanism for describing the susceptibility minimum below Tc is misleading. On the other hand, the most prominent prediction of MCT below Tc is the anomaly of the nonergodicity parameter, which, as discussed, is found by different model-independent approaches. However, within the framework of MCT, this anomaly is closely connected with the appearance of a so-called knee feature in the spectral shape of the fast dynamics spectrum below Tc. This feature, however, has not been identified experimentally in molecular liquids, and only indications for its existence are observed in colloidal systems [19]. In molecular systems, merely a more or less smooth crossover to a white noise spectrum has been reported in some cases [183,231,401]. Thus, it may be possible that the knee phenomenon is also smeared out. 

The simplest prominence prediction algorithm simply uses the concatenation of the lexical prominence patterns of the words. So for ... 

Prominence prediction by deterministic means is actually one of the most successful uses of non-statistical methods in speech synthesis. This can be attributed to a number of factors, for example the fact that the rules often don t interact or the fact that many of the rules are base on semantic features (such that even if we did use a data driven technique we would still have to come up with the semantic taxonomy by hand). Sproat notes [410] that statistical approaches have had only limited success as the issue (especially in compound noun phrases) is really one of breadth and not modelling regardless of how the prominence algorithm actually works, what it requires is a broad and exhaustive list of examples of compound nouns. Few complex generalisations are present (what machine learning algorithms are good at) and once presented with an example, the rules are not difficult to write by hand. 

Phrasing and prominence prediction can be performed by a variety of algorithms, including simple rules, sophisticated rules, data driven techniques and statistical techniques. 

The Fujisaki model is most commonly used with Japanese, but has been used or adapted to many other languages. In Japanese, we find that the range of pitch accent phenomena is narrower than in languages such as English, which means that the model s single type of accent is particularly suited. In addition, the nature of intonation in Japanese means that accents are marked in the lexicon, which greatly simplifies the problem of prominence prediction. Hence a simple approach to this, which uses accent information from the lexicon alone, is often sufficient. A common approach therefore is to determine phrase breaks and prominent syllables from the text, and then phrase by phrase and syllable by syllable generate the input command parameters for the Fujisaki... 

---