Offre en lien avec l’Action/le Réseau : – — –/– — –

Laboratoire/Entreprise : ICube, Strasbourg
Durée : 3 ans
Contact : jonathan.chardin@cephalgo.com
Date limite de publication : 2022-11-30

Contexte :
The thesis will be carried out in partnership between the company CEPHALGO (specialised in the development of hardware for the recording of EEG signals and their statistical study) under the supervision of Dr Jonathan Chardin and the ICube research laboratory under the supervision of Dr Thomas Lampert HDR, Chair of Data Science and AI (specialised in the development of machine learning and deep learning models). The candidate will share his/her time between the CEPHALGO company and the ICube research laboratory.

Keywords: Deep learning, Affective computing, Valence arousal model, Fourier transform, classifiers, machine learning, PCA, wavelets, statistical analysis.

Sujet :
We are looking for a candidate for a PhD thesis in the field of emotion recognition from electroencephalographic (EEG) signals. The objective of the thesis is to perform statistical analysis of EEG signals and to find correlations between quantities extracted from the raw signals and the emotions associated with the signals during their recording. The candidate will first acquire EEG signals from several users to build a study database. The thesis will then have two stages:
1. As a comparison method, a ‘classical’ analysis of the data will be performed. That is to perform any necessary pre-processing (filtering, elimination of bad signals, estimation of the quality of the signals and decomposition into brain waves, i.e. alpha, beta, theta, gamma waves…), make a statistical study on the data (extraction of quantities from the processed signals, i.e. Hjorth parameters, spectral entropy, moments…), and to find correlations between these different quantities and the emotions associated with the signals.
2. This will lead to the development of machine learning ( particularly deep learning) algorithms to associate EEG signals with their corresponding emotions based on state-of-the-art models (transformers, convolutional neural networks, etc).

Profil du candidat :
Desirable skills:
– Good knowledge of signal processing (Fourier transform, wavelet decomposition, spectrograms, etc.)
– Experience in working with EEG data or time-series would be a plus but not necessary
– Care for patients suffering from mental disorders

Formation et compétences requises :
Skills required:
– Master’s degree (M2) in Computer Science or similar with a strong mathematical component
– Experience in machine learning projects, preferably in addition to Deep Learning
– A solid knowledge of the python programming language and associated libraries (numpy, scipy, matplotlib)
– Project management skills when collaborating with other research partners
– Good interpersonal skills to interact with medical professionals and patients
– Adventurous towards the dynamic startup environment and scientific challenges

Adresse d’emploi :
CEPHALGO, Strasbourg