EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Reconstructing voice identity from noninvasive auditory cortex recordings

Charly Lamothe iD, Etienne Thoret iD, Régis Trapeau iD, Bruno L Giordano iD, Julien Sein iD, Sylvain Takerkart iD, Stephane Ayache iD, Thierry Artieres iD, Pascal Belin iD

eLife · 2026

DOI: 10.7554/elife.98047 ↗

Kurz gefasst

Fachartikel in der Zeitschrift eLife, 2026. Verfasst von Charly Lamothe und 8 weiteren Autor:innen. Beteiligte Einrichtung: La Timone Neuroscience Institute UMR 7289, CNRS, Aix-Marseille University, Marseille, France. Hauptthemen laut MEDLINE: Auditory Cortex, Voice. Die Arbeit nennt 92 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Research Council, France Life Imaging network, Institute of Convergence ILCB.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The cerebral processing of voice information is known to engage, in human as well as non-human primates, ‘temporal voice areas’ (TVAs) that respond preferentially to conspecific vocalizations. However, how voice information is represented by neuronal populations in these areas, particularly speaker identity information, remains poorly understood. Here, we used a deep neural network (DNN) to generate a high-level, small-dimension representational space for voice identity—the ‘voice latent space’ (VLS)—and examined its linear relation with cerebral activity via encoding, representational similarity, and decoding analyses. We find that the VLS maps onto fMRI measures of cerebral activity in response to tens of thousands of voice stimuli from hundreds of different speaker identities and better accounts for the representational geometry for speaker identity in the TVAs than in A1. Moreover, the VLS allowed TVA-based reconstructions of voice stimuli that preserved essential aspects of speaker identity as assessed by both machine classifiers and human listeners. These results indicate that the DNN-derived VLS provides high-level representations of voice identity information in the TVAs.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Auditory CortexVoiceHumansMagnetic Resonance ImagingFemaleMaleAcoustic StimulationNeural Networks, ComputerBrain MappingAdultYoung Adult

Stichwörter der Autor:innen

DNNscomputational biologyfMRIhumanneurosciencesystems biologytemporal voice areasvoice perceptionvoice reconstruction

Autor:innen und Institutionen

  • La Timone Neuroscience Institute UMR 7289, CNRS, Aix-Marseille University, Marseille, France.
  • Laboratoire d'Informatique et Systèmes UMR 7020, CNRS, Aix-Marseille University, Marseille, France.
  • Perception, Representation, Image, Sound, Music UMR 7061, CNRS, Marseille, France.
  • Institute of Language Communication & the Brain, Marseille, France.
  • Centre IRM-INT@CERIMED, Marseille, France.
  • École Centrale de Marseille, Marseille, France.

ORCID-Profile: Charly Lamothe, Etienne Thoret, Régis Trapeau, Bruno L Giordano, Julien Sein, Sylvain Takerkart, Stephane Ayache, Thierry Artieres, Pascal Belin

Förderung

  • European Research Council
  • France Life Imaging network
  • Institute of Convergence ILCB

Zitieren

APA

Lamothe, C., Thoret, E., Trapeau, R., Giordano, B. L., Sein, J., Takerkart, S., Ayache, S., Artieres, T., & Belin, P. (2026). Reconstructing voice identity from noninvasive auditory cortex recordings. eLife. https://doi.org/10.7554/elife.98047

Vancouver

Lamothe C, Thoret E, Trapeau R, Giordano BL, Sein J, Takerkart S, et al. Reconstructing voice identity from noninvasive auditory cortex recordings. eLife. 2026. doi:10.7554/elife.98047

BibTeX
@article{Lamothe2026,
  author = {Lamothe, Charly and Thoret, Etienne and Trapeau, Régis and Giordano, Bruno L and Sein, Julien and Takerkart, Sylvain and Ayache, Stephane and Artieres, Thierry and Belin, Pascal},
  title = {Reconstructing voice identity from noninvasive auditory cortex recordings},
  journal = {eLife},
  year = {2026},
  doi = {10.7554/elife.98047}
}
Kostenlos starten →