Effect of repeated listening experiences on the intelligibility of synthesized speech

Abstract

Two studies (one with child subjects, one with adult subjects) were conducted to investigate the effect of the following variables on the intelligibility of synthesized speech: synthesizer (DECtalk child's voice vs. Echo II+); repeated listening experiences (five sessions); and vocabulary type (novel vs. repeated). The pattern of findings was similar for the two studies: intelligibility scores were significantly higher for the DECtalk than for the Echo; repeated listening experiences resulted in significantly improved scores for both novel and repeated vocabulary; and, in sessions 2 to 5, repeated vocabulary was more intelligible than novel vocabulary. The results provide evidence that repeated listening experiences result in improved performance for both children and adults. Children and adults not only remember and recognize words that they have heard previously (repeated vocabulary), but repeated exposure to synthesized speech also improves performance on words heard for the first time (novel vocabulary).

• adults
• children
• intelligibility
• learning theory
• speech synthesis