Gaétan Lepage

Gaétan Lepage

Welcome to my personal web page !

🎓 Curriculum vitae

Here is my CV: cv.pdf

Education

[Oct. 2020 - now ] - PhD Student (INRIA - Perception team, Grenoble, FRANCE)

Deep Probabilistic Reinforcement Learning for Audio-Visual Human-Robot Interaction.
Supervisors: Xavier Alameda-Pineda, Laurent Girin and Chris Reinke


[2019 - 2020] - Master of Science (Université Grenoble Alpes, Grenoble, FRANCE)

Master of Science in Industrial and Applied Mathematics.
Track: Data science.


[2017 - 2020] - Master of Engineering / Diplôme d’Ingénieur (Grenoble INP - Ensimag, Grenoble, FRANCE)

National School of Computer Science and Applied Mathematics.
Track: Mathematical Modeling, Image and Simulation.


[2015 - 2017] - Classe Préparatoire aux Grandes Ecoles (Lycée La Martinière Monplaisir, Lyon, FRANCE)

Mathematics and Physics courses in leading French University preparatory program for scientific studies highly competitive entry exams.
Track: PTSI - PT*


Experience

[Apr. - Aug. 2020] - Research Intern (HES-SO Valais-Wallis - MedGIFT team, Sierre, Valais, SWITZERLAND)

I worked under the supervision of Dr. Henning Müller and Dr. Manfredo Atzori (MedGIFT team) on a computer vision problem with medical applications. The task is part of the european ExaMode project.

My goal was to handle the segmentation of compound figures (figures involving multiple sub-figures) from medical publications. To achieve this, I investigated deep learning state of the art models and adapted them to this problem.

Supervisors: Dr. Henning Müller and Dr. Manfredo Atzori
Software: https://github.com/GaetanLepage/compound-figure-separator


[Feb. 2018 - Feb. 2020] - BugBuster (Grenoble INP - Ensimag, Grenoble FRANCE)

Member of the IT department in my engineering school. I was helping other CS students with their personal computer. This job involved installing OS and teaching students about Linux. Troubleshooting and communication were two ubiquitous and essential skills.


[Jun. - Sep. 2019] - Intern (Hoomano, Lyon FRANCE)

Hoomano (social robotics startup), 3.5 months internship
Subject: Development of human-robots interactions.
Languages: Python, (Java, Dart/Flutter)
I worked on two very interesting topics during this internship :

  • On the one hand I made some R&D work about the potential use of audio signal in social robotics. Indeed, image analysis is the only source of data currently used to infer the user’s level of engagement in human-robots interactions. After studying the state-of-the-art I got to develop and integrate some feature extraction tools to enhance the existing solution.

  • On the other hand I had the opportunity to get involve in one of the company’s customer projects. I developed an interaction engine that managed the interactions with a human-like avatar on a tablet. The product aimed at providing pleasant interactions to elderly. The first real-life tests were very encouraging.