Description du sujet de thèse

As part of the ERC-Synergy NASCE project (“Natural Auditory SCEnes in Humans and Machines”), this PhD aims to understand how the human brain processes real-world auditory scenes (environmental sounds, multi-source mixtures) and identify the mechanisms enabling their analysis and interpretation.
The PhD student will focus specifically on:
• Setting up behavioral experiments combining laboratory tests and large-scale online (crowdsourcing) studies.
• Acquiring and analyzing MEG (Magnetoencephalography) data to study the temporal dynamics of auditory perception, particularly the influence of low-level (acoustic) and high-level (semantic, attentional) factors.
• Employing advanced analyses (e.g., representational similarity analysis, connectivity analysis) to compare brain responses with predictions of computational models (deep neural networks developed by the NASCE team).
The objectives include assessing how the brain segments, groups, or highlights certain sound sources in complex scenes, and determining how these processes are modulated by context (noise, multiple sources, attention).

Contexte de travail

The PhD student will be hosted at the Institut de Neurosciences de la Timone (INT), a Joint Research Unit (UMR) under the supervision of the CNRS and Aix-Marseille University, located on the Timone campus.
This thesis is part of the ERC-Synergy NASCE project, coordinated by Dr. Bruno Giordano (CNRS, Marseille) and Prof. Elia Formisano (Maastricht University, The Netherlands). NASCE is based on an interdisciplinary collaboration (neuroscience, artificial intelligence, cognitive psychology) and aims to develop a novel theory of auditory scene analysis in the brain and in machines. In Marseille, the group focuses on electrophysiological approaches (MEG, iEEG) and signal processing, while in Maastricht, the partner team provides ultra-high-field imaging (7T and 9.4T fMRI) and AI-based modeling.
The PhD student will be enrolled at the Aix-Marseille University Doctoral School (NeuroSchool program). Regular interactions with team members in Maastricht are planned (videoconference meetings, joint workshops).
Candidate Profile
• Required degree: Master's (or equivalent) in neuroscience, psychology, biomedical engineering, computer science, or related discipline.
• Skills: Python programming, solid knowledge of statistics, behavioral protocol design and analysis. Prior experience with MEG or EEG is a plus. Good command of scientific English (oral/written).
• Qualities: Autonomy, appetite for interdisciplinary work, organizational skills, and team spirit.

Contraintes et risques

The PhD project does not involve any particular constraints or risks.