Missions

Within the ERC-Synergy project NASCE (“Natural Auditory SCEnes in Humans and Machines”), the post-doctoral fellow will:
• analyse MEG and iEEG data acquired while participants listen to natural sounds;
• compare these signals with lightweight, brain-inspired deep neural networks;
• ensure methodological alignment and data exchange between INT-Marseille (electrophysiology) and Maastricht University (ultra-high-field MRI & AI).

Activités

• Time-frequency and multivariate analyses (RSA, encoding models) using Python / MNE.
• Training and evaluation of convolutional-recurrent networks on ecological sound corpora.
• Writing manuscripts, preparing presentations, and openly sharing data and code.

Compétences

• PhD in neuroscience, cognitive science, biomedical engineering, or a related field, with a focus on audition and/or computational modelling.
• Experience in multivariate brain-data analysis and, ideally, brain-model alignment with deep neural networks.
• Proficiency in Python for deep learning (NumPy, PyTorch).
• Competence with representational similarity / encoding models and, where necessary, functional connectivity (coherence, causality, etc.).
• Strong motivation for interdisciplinary teamwork and communication in an international environment.
• Scientific rigor, organisational skills, and the ability to prioritise within a multi-site collaboration.
• Excellent written and spoken English; knowledge of French is an asset but not mandatory.

Contexte de travail

The position is hosted at the Institut de Neurosciences de la Timone (INT), a CNRS / Aix-Marseille University laboratory on the Timone campus in Marseille. You will work in the team of Dr Bruno Giordano.
The NASCE project combines expertise in neuroscience, artificial intelligence, and cognitive psychology to develop a unified theory of auditory-scene analysis in brains and machines. Close collaboration with Prof Elia Formisano's group (Maastricht University) will give you direct access to the ultra-high-field MRI and AI-modelling work packages.

Contraintes et risques

No constraints or risks