Segregation of vowel sequences by normalhearing and hearing impaired listeners

1
Segregation of vowel sequencesby normal-hearing
andhearing impaired listeners

• Étienne Gaudrain1, Nicolas Grimault1, Eric W.
  Healy2, Jean-Christophe Béra3
• IAPA 2005

1. Laboratoire de Neurosciences et Systèmes
Sensoriels,CNRS UMR 5020,Université Claude
Bernard Lyon 1,Lyon, France
2. Speech Psychoacoustics Laboratory, Department
of Communication Sciences and Disorders,Universit
y of South Carolina,Columbia, 29208 USA
3. Inserm UMR 556,Lyon, France
2
Context

• Auditory Scene Analysis (BREGMAN, 1990)
• Segregation Streaming ?
  acoustic sources
• Mechanisms

Simultaneous
Sequential
t
t
Voice 1
t
Voice 2
t
3
Previous studies

• Stimuli
• Pure tones e.g. van Noorden, 1975
• Complex tonese.g. Vliegen and Oxenham, 1999
  Grimault et al., 2001

4
Previous studies

• Conclusions
• Pitch height is a factor of sequential
  segregation
• 2 kinds of cues can be useful for segregation
• Spectral cues (tonotopy)
• Temporal cues
• The HI perceive less spectral cues than NH, but
  as much temporal cues, which are used for
  segregation

5
Object

• Are these results unchanged for speech stimuli ?
• How important are each kind of cue for sequential
  segregation ?

6
Experiment
110 Hz
a é i
o u ou
100 Hz
238 Hz
i a u
o ou é
100 Hz
7
Subjects

• Normal hearing
• 14 young (YNH), mean age 23
• 4 elderly (ENH), mean age 49
• Hearing impaired
• 5 young (YHI), mean age 21
• 6 elderly (EHI), mean age 67

8
Results - 1
1 stream
Mean results across subjects and repetitions
2 streams
9
Conclusion

• Strong age effect
• Further testing for YHI is requested, with
  perfect recognition conditions

10
Results - 2
Elderly Hearing Impaired subjects with and
without hearing aid
11
Acknowledgment

• Entendre Hearing Aid Dispenser Group
• Région Rhône-Alpes