When a talker is instructed to speak clearly, the resulting """"""""clear speech"""""""" is easier to understand than typical conversational speech. In addition, clear and conversational speech differ on many acoustic dimensions. Recent work capitalizing on talker differences in clear speech intelligibility and acoustic effects suggests that increased vowel duration and an expanded vowel space lead to improved vowel intelligibility for listeners with normal hearing. However, it is not clear whether these acoustic changes are actually beneficial for listeners with hearing impairment, whether applying these changes via signal processing leads to improved vowel intelligibility, or whether the acoustic changes that enhance vowel intelligibility also make words easier to understand. The four proposed studies will use materials selected from the Ferguson clear speech database (2004), which includes conversational and clear speech produced by 41 talkers. The first proposed study extends the PI's earlier work to the clinical population for which clear speech is intended (i.e., listeners with hearing loss), measuring vowel intelligibility in this population for all of the talkers in the database. In the second study, vowel acoustic characteristics will be measured in clear and conversational speech for all 41 talkers. When the perceptual and acoustic data from these two studies are combined, along with previous perceptual data from listeners with normal hearing, a wide range of statistical techniques can be used to explore which specific acoustic changes lead to improved vowel intelligibility for listeners with normal and impaired hearing. The third and fourth studies move this line of research into some important new directions. The third study investigates new techniques by which hearing aid algorithms might enhance the intelligibility of conversational speech, while the fourth study begins to investigate more complex speech materials, namely monosyllabic words. By revealing the specific acoustic characteristics that are important for vowel intelligibility in listeners with hearing loss, the results of these experiments will contribute to our overall understanding of the mechanisms underlying speech perception in normal and impaired listeners. In addition, the results will suggest new hearing aid algorithms as well as strategies for helping the frequent communication partners of listeners with hearing loss to speak more clearly. Finally, the talker differences revealed in these studies have implications for models of speech production, in particular Lindblom's H&H Theory (1990). When a talker is asked to speak clearly, the resulting """"""""clear speech"""""""" is usually easier for listeners with hearing loss to understand than, and also differs acoustically from, typical conversational speech. The proposed research will seek to determine exactly which clear speech acoustic changes are actually responsible for making clear speech easier to understand. This will help in developing new hearing aid algorithms as well as new techniques for training the communication partners of individuals with hearing loss.
|Ferguson, Sarah Hargus; Quené, Hugo (2014) Acoustic correlates of vowel intelligibility in clear and conversational speech for young normal-hearing and elderly hearing-impaired listeners. J Acoust Soc Am 135:3570-84|
|Ferguson, Sarah Hargus (2012) Talker differences in clear and conversational speech: vowel intelligibility for older adults with hearing loss. J Speech Lang Hear Res 55:779-90|
|Bent, Tessa; Kewley-Port, Diane; Ferguson, Sarah Hargus (2010) Across-talker effects on non-native listeners' vowel perception in noise. J Acoust Soc Am 128:3142-51|