The influence of talker differences on vowel identification by normal-hearing and hearing-impaired listeners

Abstract

Vowel identification was tested in quiet, noise, and reverberation with 20 normal-hearing subjects and 20 hearing-impaired subjects. Stimuli were 15 English vowels spoken in a /b-t/context by six male talkers. Each talker produced five tokens of each vowel. In quiet, all stimuli were identified by two judges as the intended targets. The stimuli were degraded by reverberation or speech-spectrum noise. Vowel identification scores depended upon talker, listening condition, and subject type. The relationship between identification errors and spectral details of the vowels is discussed.

Audibility of American English vowels produced by English-, Chinese-, and Korean-native speakers in long-term speech-shaped noise

“…Kiefte et al (2010) showed that identification of /u/ and /i/ in quiet was significantly affected by formant amplitude for steady vowels. Nábélék et al (1992) reported that listeners sometimes misidentified /u/ as /i/ mainly due to the noise masking the F2 of /u/. The findings of the present and previous studies of vowel detection using noise indicate that /i/ is the most audible while /u/ is the least audible, regardless of speaker.…”

“…Thus, detection thresholds were lower for the front vowels than for the back vowels. An influence of formant amplitude on vowel perception was also found by Nábélék et al (1992) and Kiefte et al (2010). Kiefte et al (2010) showed that identification of /u/ and /i/ in quiet was significantly affected by formant amplitude for steady vowels.…”

Evaluating models of vowel perception

“…Assmann and Summerfield ͑1989͒, for example, reported that listeners can accurately identify concurrent vowels, even when the individual formants have considerable overlap in the spectrum. For normally hearing listeners, vowel identification accuracy is fairly robust in noise ͑Nábělek et al, 1992͒ andwith reduced spectral resolution ͑Dubno andDorman, 1987;Fu et al, 1998͒. Additionally, although cochlear hearing impairment results in broadening of auditory filters and loss of spectral contrast, in the absence of background noise, vowel identification performance can remain high for listeners with mild to moderate losses ͑Nábělek et al., 1992;Owens et al, 1968;Van Tasell et al, 1987͒. A successful model of vowel perception will need to account for the levels of performance observed in each of these less-than-ideal situations.…”

The relative importance of spectral cues for vowel recognition in severe noise

“…Previous studies have investigated vowel cues in noise for SNRs of À5 dB and higher (Nabelek et al, 1992;Parikh and Loizou, 2005). The latter authors reported vowel recognition scores as high as 75% (in a set containing 11 vowels) at SNR levels of À5 dB.…”

“…Poor speech intelligibility in noise is due to the masking of information that would normally be available to listeners, leaving them with access to only certain parts of the spectral and temporal information in the signal (Juang, 1991;Nabelek et al, 1992). The present study considered vowel recognition in noise, and specifically investigated the importance of formants when listening in severe noise.…”