Voice recognition in brain

Voice recognition in brain

1. BRAIN AREAS

   • temporal voice areas
     • process acoustical info regardless of voice fam
       • Latinus et al 2011
     • Belin 2000
   • Right inf frontal areas
     • learned voices perceptually identical
     • = voice identity processing

2. MODELS

   • HIerarchical processing
     • Belin 2011
     • 1) acoustic analysis
     • 2) voice compared to long-term voice representations
     • respond to fam voices independent of speech ctxt
   • PROTOTYPE
     • comparaison d'une voix entendue avec un prototype [voir Kahn]
       • plus ou moins proche du prototype
       • plus proche = plus dur à distinguer
       • - besoin du prototype pour voix connues
     • prototype construit à partir des voix connues = langue
     • comme pour visages
       • Valentine 95, Sangrigoli 2005 [Kahn]
     • Lavner 2001
     • Papcun 1989
   • PARAMETRES ACOUSTIQUES
     • Bricker et al 76 [Kah]
     • extraction d'indices acoustiques, comparaison avec paroles connues
       • comparaison avec memoire
       • = gd rôle fréq voix
     • parce que plus rapide ID phonemes que locuteurs
       • Cutler 2011 [Kahn]

3. FACE RECOGNITION

   • N170 (occipitotemporal)
     • face encoding
     • Eimer 2011
   • N250 (occipitotemporal)
     • acquisition & activation of representations for indvidual faces
       • Kaufmann et al 2009
       • Zimmermann Eimer 2013
   • centroparietal pos (300ms)
     • retrieval of episodic/semantic person info
       • Schweinberger Burton 2003
       • Wiese 2012
     • also for OLD/NEW effects
       • Friedman Johnson 2000
   • Dm = Differences due to subsequent memory
     • face encoding into memory
       • Sommer et al 1991

4. AUDITORY RECOGNITION

   • possible candidates
     • N1-P2
       • sensitive to sound repetitions (incl.voices)
         • Schweinberger 2001, 2011
     • 160 ms: frontotemporal positivity to voice
       • Charest et al 2009
       • discrim voices vs sounds
     • Frontocentral MMN, 200ms
       • detection of familiar among unfamiliar voices (pas sur)
         • Beauchemin et al 2006
     • parietal P3
       • sensitive to voice repetitions despite phonetic changes
         • Schweinberger et al 2011
   • Voice learning
     • ERP Dm: pos ~250ms, suistained, right parietal
       • = successful encoding
       • for voices subsequently recognized
       • Zaske 2014
     • 300-700ms ERP Old/New effect
       • only when same sentences as in study
       • = retrieval of study episode
       • Zaske 2014
     • power differences in beta-band oscillations (16-17 Hz), 290-370ms, central & right temp
       • voice identity
       • formation of cat-specific representations in long-term memory
       • Zaske 2014
     • reduction of N1-P2 compex + reduction of P3
       • Schweinberger 2011
       • = adaptation
     • voice identity processed in auditory cortex ~250ms after onset
       • + less speech-content dependent after 300ms
         • Schweinberger 2011