2021 年 26 巻 1 号 p. 32-41
It has been reported that communication difficulties caused by hearing loss can lead to depression and cognitive decline. There are various ways to improve audibility by processing the original voice and presenting it to the listeners. This research aims to develop an intelligent hearing aid that has not only the ability to amplify and process the outer sound waves but also the ability to modify the speed of speech. In this paper, we designed two methods for improving audibility by speech rate conversion in real time: CE (Consonant Elongation) system and FE (Forward Elongation) system. In the CE system, only the consonant part of the word is elongated, whereas in the FE system, only the first half of the word is elongated. Both methods are designed to prevent the word lengths from being elongated excessively. We evaluated these two methods and found that the percentage of correct responses using the FE system was higher than that of the original voice.