Diane Kewley-Port

The Indiana Speech Training Aid (ISTRA) II: Training curriculum and selected case studies

A new approach to computer-based speech training is described. The ISTRA system is being developed as a training aid, intended for use by speech-language pathologists and teachers of the deaf. The system consists of a microcomputer equipped with a low-cost, speaker-dependent speech recognizer, and a series of programs for use in various phases of speech training. On-line feedback is evaluative and reflects the quality of whole words, based on the level of agreement between stored templates and new utterances. A variety of software to conduct speech drills has been developed, which includes several modes of training recommended for use in a training curriculum. Clients are able to run their speech drills independently. Three case studies are presented to demonstrate the use of ISTRA in speech training with hearing-impaired and normal-hearing, misarticulating children. Improvement of the quality of words trained, as well as generalization to non-trained words, in some cases, was demonstrated.

Factors affecting vowel formant discrimination by hearing-impaired listeners.

The goal of this study was to measure the ability of adult hearing-impaired listeners to discriminate formant frequency for vowels in isolation, syllables, and sentences. Vowel formant discrimination for F1 and F2 for the vowels /I epsilon ae / was measured. Four experimental factors were manipulated including linguistic context (isolated vowels, syllables, and sentences), signal level (70 and 95 dB SPL), formant frequency, and cognitive load. A complex identification task was added to the formant discrimination task only for sentences to assess effects of cognitive load. Results showed significant elevation in formant thresholds as formant frequency and linguistic context increased. Higher signal level also elevated formant thresholds primarily for F2. However, no effect of the additional identification task on the formant discrimination was observed. In comparable conditions, these hearing-impaired listeners had elevated thresholds for formant discrimination compared to young normal-hearing listeners primarily for F2. Altogether, poorer performance for formant discrimination for these adult hearing-impaired listeners was mainly caused by hearing loss rather than cognitive difficulty for tasks implemented in this study.

Second language vowel training using vowel subsets: Order of training and choice of contrasts

Our previous vowel training study for Japanese learners of American English [J. Acoust. Soc. Am. 117, 2401 (2005)] compared training for two vowel subsets: nine vowels covering the entire vowel space (9V condition); and the three more difficult vowels (3V condition). Trainees in 9V condition improved on all vowels, but their identification of the three more difficult vowels was lower than that of 3V trainees. Trainees in 3V condition improved identification of the trained three vowels but not the other vowels. In order to further explore more effective training protocols, the present study compared two groups of native Korean trainees using two different training orders for the two vowel subsets: 3V then 9V (3V‐9V); and 9V then 3V (9V‐3V). The groups were compared in terms of their performance on all nine vowels for pre‐, mid‐, and post‐test scores. Average test scores across the two groups were not different from each other. A closer examination indicated that group 3V‐9V did not improve on one of the th...

Small Business Innovation Research (SBIR) funding: A case study in bringing a computer‐based speech training aid into the marketplace

University research projects occasionally involve applications of speech technology that are intended to produce a commercial product, such as the development of a microcomputer‐based speech training aid (ISTRA) at Indiana University. Even though initially funded as basic research, such projects, when successful, eventually require an alternative source of funds. SBIR programs in most federal agencies have been mandated by Congress to support research and development in small businesses. This presentation will describe SBIR programs and the funding process, from the perspective of investigators who attempt to form small businesses outside of the university womb. [Research support by NSF and NIH, SBIR Phase I grants.]

Evaluation of a speaker‐dependent recognition metric as a substitute for human judgments of speech quality

The Indiana Speech Training Aid project (ISTRA) is evaluating the use of speaker‐dependent speech recognition to provide speech‐drill feed‐back for deaf or misarticulating children. The ISTRA systems employ IBM PCs with Interstate Voice Products Vocalink SRB boards. The basic approach is to form templates from the childs best current productions. Feedback in drill sessions is then based on a goodness metric. That metric represents the “match” between 10‐ms sampling of a new utterance and the corresponding information for the stored template. The goodness metric has been partially validated as a substitute for feedback judgments provided by a human teacher. In general, the method compares values of the goodness metric to actual ratings by a jury of listeners. For words spoken by either normal speakers who intentionally vary the quality of their pronounciations, or by hearing‐impaired speakers, average inter‐listener correlations for ratings of quality did not differ from the correlations between the good...

The Indiana Speech Training Aid (ISTRA): First year report

The ISTRA systems [Watson et al., J. Acoust. Soc. Am. Suppl. 1 81, Abstract OO14 (1987)] are used to improve the speech of deaf or misarticulating children, by providing unidimensional feedback based on acoustic information distributed over whole words or other utterances. This approach differs from that of many previous aids in which feedback was based on measures of single acoustic dimensions, or others that provided spectrographic displays. Using software developed over the first year of the project, the clinical effectiveness of ISRA drill has been evaluated for several deaf and one normal‐hearing, misarticulating child. The research emphasis has not been on comparisons between the progress of training with a teacher and that with the ISTRA system, but rather on demonstrations that ISTRA drill is itself an effective training method. In these studies, listener juries rate the speech of words collected before, during, and after training. Results show significant improvement of the quality of words train...