Session Speaker Charact.:Speaker Characteristics
|
| Speaker Charact.-1 |
HIERARCHICAL CLUSTERING OF SPEAKERS INTO ACCENTS WITH THE ACCDIST METRIC
Mark Huckvale, University College London Paper File |
| Hierarchical clustering of speakers by their pronunciation patterns could be a useful technique for the discovery of accents and the relationships between accents and sociological variables. However it is first necessary to ensure that the clustering is not influenced by the physical characteristics of the speakers. In this study a number of approaches to agglomerative hierarchical clustering of 275 speakers from 14 regional accent groups of the British Isles are formally evaluated. The ACCDIST metric is shown to have superior performance both in terms of accent purity in the cluster tree and in terms of the interpretability of the higher-levels of the tree. Although operating from robust spectral envelope features, the ACCDIST measure also showed the least sensitivity to speaker gender. The conclusion is that, if performed with care, hierarchical clustering could be a useful technique for discovery of accent groups from the bottom up. | |
| Speaker Charact.-2 |
Discrimination of Speakers Using the Formant Dynamics of /u:/ in British English
Kirsty McDougall, University of Cambridge Francis Nolan, University of Cambridge Paper File |
| Formant dynamics are an interesting source of speaker-discriminating information, reflecting both differences in speakers¡¯ vocal tract morphology and individual differences in the articulatory trajectories chosen to produce each sound. A study of speaker-distinguishing properties of the formant dynamics of /u:/ in Standard Southern British English is presented. Measurements at equidistant intervals along the F1 and F2 contours are compared with polynomial characterisations of the contours. Approximating the contours with quadratic and cubic polynomials allows more speaker-discriminating information to be conveyed with fewer parameters. The best value per predictor is provided by the cubic approximation of F2. | |
| Speaker Charact.-3 |
Phonetic content influences voice discriminability
Attila Andics, Max Planck Institute for Psycholinguistics James M. McQueen, Max Planck Institute for Psycholinguistics Miranda van Turennout, FC Donders Centre for Cognitive Neuroimaging Paper File |
| We present results from an experiment which shows that voice perception is influenced by the phonetic content of speech. Dutch listeners were presented with thirteen speakers pronouncing CVC words with systematically varying segmental content, and they had to discriminate the speakers’ voices. Results show that certain segments help listeners discriminate voices more than other segments do. Voice information can be extracted from every segmental position of a monosyllabic word and is processed rapidly, and vowel changes seem to make a greater difference than consonant changes do. We also show that although relative discriminability within a closed set of voices appears to be a stable property of a voice, it is also influenced by segmental cues – that is, perceived uniqueness of a voice depends on what that voice says. | |