Continuing evaluation of the Queen's University tactile vocoder II: Identification of open set sentences and tracking narrative

Abstract

Identification of open set words by an experienced normal hearing subject using the tactile vocoder developed at Queen's University was examined. The tactile vocoder filters and processes the acoustic waveform into 16 filter channels each of which activates a vibrator on the skin surface. After acquiring a 250-word vocabulary through the tactile vocoder, the subject was presented with 1000 different open set sentences in two reception conditions. The percentages of words correctly identified within the sentences in the lipreading (L) and lipreading plus tactile vocoder (L + TV) conditions were 57.8 percent and 79.6 percent, respectively. After two presentations of the stimulus sentences, the number of sentences repeated verbatim in the L + TV condition was 58.4 percent, which was 27.8 percent higher than scores obtained in the L condition. Performance on the tactile vocoder was also evaluated using the "tracking procedure," in which phrases of text are read to the subject who attempts to repeat the material verbatim. The final tracking rate obtained using lipreading was 15.3 words per minute, which was considerably lower than the L + TV tracking rate of 49.3 words per minute. The tactile vocoder provided valuable information which improved the subject's lipreading ability.