Auditory-Visual Speech Processing (AVSP) 2009

University of East Anglia, Norwich, UK
September 10-13, 2009

Recognizing Spoken Vowels in Multi-Talker Babble: Spectral and Visual Speech Cues

Chris Davis, Jeesun Kim

MARCS Auditory Laboratories, University of Western Sydney, Australia

It has been proposed that both spectral and visual speech cues assist in segregating a talker from noise. To test how these cues interact, the experiment examined vowel identification (in hVd context) when presented in multi-talker babble. The availability of spectral cues was manipulated by filtering the signal into (1) 8 frequency amplitude-envelope bands or (2) the same bands with additional spectral cues. The availability of visual speech cues was manipulated by using auditoryonly (AO) and auditory-visual (AV) presentations. It was found that the intelligibility benefit when spectral and visual speech cues were combined appeared to be less than that produced by adding the benefits for each cue type when tested separately. This pattern suggests that both cues provide similar information.

Full Paper

Bibliographic reference.  Davis, Chris / Kim, Jeesun (2009): "Recognizing spoken vowels in multi-talker babble: spectral and visual speech cues", In AVSP-2009, 130-133.