Open positions

We are always looking for new group members with passion, talent, and grit!

We offer you the opportunity to work in an English speaking team on exciting and fulfilling topics in the frontier between machine learning and the experimental sciences.

You’ll have a chance to work on data acquisition and analysis on the field, including the design and setup of innovative marine passive acoustic monitoring stations, the development of deep learning approaches to better understand our usage and acquisition of speech and language, and the use of advanced machine learning to improve our understanding of the universe.

Current open positions

We currently do not have any open positions, but don’t hesitate in getting in contact with us by email and we’ll figure something out.

Applications for PhD and Postdoc positions

If you are interested in working with us as a PhD student or postdoc, please send me an email. State briefly why you are interested and attach a CV, including information about the grades you had as an undergraduate. No need for a separate cover letter or certificates.

Important: please insert “Application PhD” or “Application Postdoc” in the subject line. If you are applying to a specific advertisement, note this in your email.

There are postdoc scholarship available. I’d be happy to support you after you apply to our group. Take a look at the Marie Curie fellowship.

Master projects for UTLN students

If you are a Master student at Université de Toulon looking for a Master project, contact any group member per email.

####Internship offer:

####ML-based cardiac function assessment from MRI

Context: An internship is available in the context of an international and multidisciplinary project with the Bristol Heart Institute (BHI, UK). The goal of this partnership is to improve the assessment of cardiac function using a direct assessment of heartbeat movement quality. This new measure will be fully automated to free up time for medical experts. It will be based on a modelling of myocardium deformations during heartbeat.

Planned work: During preliminary studies, the following has been achieved:

1) development of a methodology for quantifying the quality of movements: http://www.bmva.org/bmvc/2014/files/paper058.pdf

2) first step towards the application of this method to the heartbeat movement, by demonstrating that it is possible to produce a model of deformations that is suitable to serve as a basis for our quality assessment method: https://miua2018.soton.ac.uk/documents/papers/MIUA2018_026.pdf

During the internship, we will continue this work to obtain a quantification of heartbeat quality. The internship will include the following steps:

1) 3D/4D reconstruction of the heart of BHI patients from MRI which were already segmented (in 2D) at BHI.

2) Construction of a reduced representation of myocardium deformations by manifold learning following the methodology that is presented here: https://miua2018.soton.ac.uk/documents/papers/MIUA2018_026.pdf . The diagnoses for BHI patients are known, therefore it will be possible to verify that this representation makes it possible to distinguish between different pathologies.

For step 3), there is a choice between:

3a) Suppression of the 3D reconstruction step by learning a direct mapping between the MRI images and the representation of step 2: training a deep neural network as in: http://openaccess.thecvf.com/content_iccv_2015_workshops/w11/papers/Crabbe_Skeleton-Free_Body_Pose_ICCV_2015_paper.pdf .

or

3b) Construction of a model of normal heart motion according to the method of: http://www.bmva.org/bmvc/2014/files/paper058.pdf , and use of this model to compute a heartbeat quality score.

Candidate profile: This internship is mainly addressed at master level students, with background in computer science or applied mathematics.

Required skills: During this internship, the following methods will be used: Markovian modeling, manifold learning, and deep learning. The intern is not expected to be an experienced user of these techniques, but he/she must be willing to learn about them. Strong bases in mathematics and statistics will be needed for this learning. It is strongly recommended to read the articles listed above to make sure that you want to work with these methods.

Good programming skills in Python are absolutely necessary.

Contact and application: To apply, please send a CV, academic transcripts, and a motivation letter to: adeline.paiement@univ-tln.fr