13th Annual Conference of the International Speech Communication Association

Portland, OR, USA
September 9-13, 2012

Intelligibility of Speech Spoken in Noise/reverberation for Older Adults in Reverberant Environments

Nao Hodoshima (1), Takayuki Arai (2), Kiyohiro Kurisu (3)

(1) Department of Information Media Technology, Tokai University, Tokyo, Japan
(2) Department of Information and Communication Sciences, Sophia University, Tokyo, Japan
(3) TOA Corporation, Hyogo, Japan

Speech intelligibility is in general lower for older adults than young adults in reverberant environments such as train stations or airports. We aim at to make speech announcements intelligible in public spaces. Speech spoken in noise, i.e., noise-induced speech, is usually more intelligible than speech spoken in a quiet environment for young people when they are heard in noise, a phenomenon called the Lombard effect. The current study applied this effect for an input of a sound reinforcement system in public spaces. The results of the listening tests conducted by 24 older adults showed that noise/reverberation-induced speech was more intelligible than speech spoken in a quiet environment when they were in reverberant environments (reverberation time of 1.4 s and 2.4 s). The results also showed that the effect of noise/reverberationinduced speech was observed when the recording and listening condition were different. For example, different reverberation times were used between the two conditions and noise-induced speech was intelligible in reverberation. The results suggest that using noise/reverberationinduced speech as an input of a sound reinforcement system might yields higher intelligibility in public spaces.

Index Terms: older adults, speech intelligibility, Lombard effect, reverberation

Full Paper

Bibliographic reference.  Hodoshima, Nao / Arai, Takayuki / Kurisu, Kiyohiro (2012): "Intelligibility of speech spoken in noise/reverberation for older adults in reverberant environments", In INTERSPEECH-2012, 1464-1467.