Rod Nicolson

Evaluation of A Viseme-Driven Talking Head

This paper introduces a three-dimensional virtual head for use in speech tutoring applications. The system achieves audiovisual speech synthesis using viseme-driven animation and a coarticulation model, to automatically generate speech from text. The talking head was evaluated using a modified rhyme test for intelligibility. The audiovisual speech animation was found to give higher intelligibility of isolated words than acoustic speech alone.

Bridging the Gap between theory and practice: A multimedia tutorial for students of voice therapy

A computer-based tutorial for speech and language therapy students is described in which a digitized video-recording of a diagnostic interview with a voice-disordered client was dynamically linked to an on-screen transcript window. A student-centered tutorial package, which provided guidance through questions, assessment tasks, and commentary from an experienced speech and language pathologist as well as the lecturer in voice disorder, was constructed. Evaluation of the tutorial as an aid to active learning, including feedback from students and supervising clinicians, is reported and the role of this type of program as a preparation for clinical practice is discussed.

The first OSCE; does students’ experience of performing in public affect their results?

BackgroundPersonal qualities have been shown to affect students’ exam results. We studied the effect of experience, and level, of public performance in music, drama, dance, sport, and debate at the time of admission to medical school as a predictor of student achievement in their first objective structured clinical examination (OSCE).MethodsA single medical school cohort (n = 265) sitting their first clinical exam in 2011 as third year students were studied. Pre-admission statements made at the time of application were coded for their stated achievements in the level of public performance; participation in each activity was scored 0–3, where 0 was no record, 1 = leisure time activity, 2 = activity at school or local level, 3 = activity at district, regional or national level. These scores were correlated to OSCE results by linear regression and t-test. Comparison was made between the highest scoring students in each area, and students scoring zero by t-test.ResultsThere was a bell shaped distribution in public performance score in this cohort. There was no significant linear regression relationship between OSCE results and overall performance score, or between any subgroups. There was a significant difference between students with high scores in theatre, debate and vocal music areas, grouped together as verbal performance, and students scoring zero in these areas. (p < 0.05, t-test) with an effect size of 0.4.ConclusionsWe found modest effects from pre-admission experience of verbal performance on students’ scores in the OSCE examination. As these data are taken from students’ admission statements, we call into question the received wisdom that such statements are unreliable.

A talking head for speech tutoring

This work applies a viseme-driven talking head in a pronunciation training system. The aim is to create a pronunciation assistant to complement traditional methods and to assist the work of a human language tutor. Visual speech can be valuable in speech tutoring applications because vision benefits human speech perception, for three reasons as suggested by Summerfield (Summerfield, 1987): It helps speaker localization, it contains speech segmental information that supplements the audio, and it provides complementary information about the place of articulation. This study aims to elucidate the benefits of visual speech in language learning.