The present study performs an investigation on several issues concerning the automatic detection of prominences. Its aim is to offer a better understanding of the prominence phenomenon in order to be able to improve existent prominence detection systems. The study is threefold: first, the presence of hidden dynamics in the sequence of prominent and non-prominent syllables is tested by comparing results obtained with CRFs and LDCRFs. Second, the size of the context to be taken into account when determining prominence was examined and third, a new set of features was investigated. The results obtained show that LDCRFs systematically outperforms CRFs, that a context of three syllables is generally sufficient for prominence detection, that syllable length is a powerful feature to include and that new features concerning pitch movements we introduce here can substitute heuristic measures used in previous works.

Index Terms: syllabic prominence, conditional random fields