Responsibilities

ENSTA Paris - Robotics track

I am co-responsible for the robotics and Intelligent Autonomous Systems track in third year at ENSTA Paris: Description of the track

Teaching

I teach the entire INF657G course and I teach the other courses with colleagues from ENSTA Paris, Ecole Polytechnique and ONERA.

Navigation for autonomous systems

IP Paris - Master DATAAI - and Ecole Polytechnique - MScT AIAVC - X-INF657G

Drones and robots must create maps of their surroundings to plan their movement and navigate. This course presents the robotic platforms and the most common sensors (vision, Lidar, intertial units, odometry …) and the different components of navigation: control; obstacle avoidance; localization; mapping (SLAM) and trajectory planning as well as filtering (Kalman filter, particle filtering, etc.) and optimization techniques used in these fields.

Navigation for autonomous systems

ENSTA Paris - ROB312

This course presents an overview of sensors, representations and different filtering methods (Kalman filter, particle filtering) and optimization techniques used for localization and mapping (SLAM) in mobile robotics and for autonomous vehicles.

Planning and control

ENSTA Paris - ROB316

This course presents the platforms and an overview of the control, trajectory planning and task planning methods used in robotics and for autonomous vehicles.

Current students and collaborations

• Guillaume Thomas (PhD 2024-…, CEA Funding): Learning fine and dexterous manipulation through vision and kinesthetic observations. Collaboration, supervised by Mathieu Grossard Quoc Cuong Pham and Jaonary Rabarisoa.
• Mathilde Jeannin (PhD 2023-…, FARO project): Guaranteed algorithms for mobile robotics. Collaboration, supervised by Eric Goubault and Sylvie Putot.
• Marwane Hariat (PhD 2022-…, FARO project): Self-supervised depth estimation for robot navigation. Collaboration, supervised by Antoine Manzanera.
• Ivelina Stoyanova (PhD 2022-…, funding THALES): Characterization of a Markov decision process in a partially observable, incremental and event-driven environment for reinforcement learning using a decision policy based on temporal rules. Co-supervised with Sao Mai Nguyen and Nicolas Museux.
• Adrien Lafage (PhD 2022-…, funding Renault): Comprehensive intermediate and long-term driving risk prediction with Deep Neural Networks. Co-supervised with Gianni Franchi and Mathieu Barbier.
• Gabriel Quere (PhD 2021-…, funding DLR) : Learning and designing shared control skills from demonstrations for assistive robots. Co-supervised with Freek Stulp.
• Tom Dupuis (PhD 2021-…, funding CEA): Reinforcement learning and representation learning for visuomotor control. Co-supervised with Jaonary Rabarisoa and Quoc Cuong Pham.
• Kai Zhang (PhD 2021-…, funding CEA): Model of the situation and task planning of a mobile manipulator in an uncertain logistical environment. Co-supervised with Julien Alexandre Dit Sandretto, Eric Lucet, Selma Kchir.

Projects

• ANR MOBILEX TAURUS (2023-2027): Participation in the MOBILEX challenge on navigation in unstructured environments. Led by Alexandre Chapoutot, in collaboration with EXAIL Robotics.
• ANR ASTRID Nimble x ENSTA (2023-2024): Extending the Aru robot’s autonomous outdoor navigation capabilities through traversability prediction - Collaboration with Nimble One
• AID FARO (2021-2025): The FARO project focuses on the development of new algorithms (model and/or data driven) and new validation methods to ensure the proper functioning of swarm drones, on land, in the air or at sea - Collaboration with Telecom Paris and Ecole Polytechnique

Former students

• David Brellmann (PhD 2020-2024): Experimental and Theoretical Analysis of Reinforcement Learning Algorithms. Co-supervised with Goran Frehse.
• Pavan R Vasishta (Postdoc 2021-2022): State representation learning for industrial applications (VeriDREAM project).
• Thomas Rojat (PhD 2020-2022): Interpretability of latent representations learned by neural networks for the characterization of abnormal events during vehicle driving. CIFRE with Renault. Co-supervised with Rodolphe Gélin, Raphael Puget & Natalia Diaz Rodriguez.
• Vyshakh Palli Thazha (PhD 2017-2022): Using context-cues and interaction for traffic-agent trajectory prediction. CIFRE with Renault. Co-supervised with Hervé Illy & Javier Ibanez-Guzman.
• Hugo Caselles-Dupré (PhD 2018-2021): On the role of Actions and Machine Learning in Artificial Agent Perception. CIFRE with Softbank Robotics, Co-supervised with Mickael Garcia-Ortiz
• Florence Carton (PhD 2017-2021, funding CEA): Exploration of reinforcement learning algorithms for autonomous vehicle visual perception and control co-supervised with Jaonary Rabarisoa and Quoc Cuong Pham
• Timothée Lesort (PhD 2017-2021): Continual Learning : Tackling Catastrophic Forgetting in Deep Neural Networks with Replay Processes. CIFRE with THALES. Co-supervised with Jean-Francois Goudou and Andrei Stoian.
• Clément Pinard (PhD 2016-2019, CIFRE Parrot): Robust Learning of a depth map for obstacle avoidance with a monocular stabilized flying camera, Co-supervised with Antoine Manzanera and Laure Chevalley.
• José Magno Mendes Filho (PhD 2015-2019): Online Distributed Motion Planning for Mobile Multi-robot Systems. Co-supervised with Eric Lucet (CEA)
• Adrien Matricon (PhD 2014-2018) : Merging robotic skills into more general skills. Co-supervised with Pierre-Yves Oudeyer.
• Harmish Khambaita (Post-Doc 2017-2018) : Decision making for autonomous vehicles
• Tianming Lu (Post-Doc, 2017-2018): Teleoperation of autonomous vehicles
• Joris Guery (PhD 2014-2017): Robust visual recognition by neural networks in robotic exploration scenarios. Detect me if you can!. Co-supervised with Bertrand Le Saulx (ONERA)
• Clément Masson (PhD 2015-2016): Unsupervised learning of sensorimotor contingencies. Co-supervised with Freek Stulp and Olivier Sigaud (ISIR).
• Panagiotis Papadakis (Post-Doc 2014-2016): Object detection and semantic mapping for a humanoid robot. As part of the ROMEO 2 project.
• Celine Craye (PhD 2014-2017): Intrinsic motivation mecanisms for incremental learning of visual saliency. Co-supervised with Jean-Francois Goudou.
• Yuxin Chen (PhD 2013-2017) : Interactive learning of words and objects for a humanoid robot.
• Louis-Charles Caron (PhD 2012-2015) : Combining Depth, Color and Position Information for Object Instance Recognition on an Indoor Mobile Robot Co-supervised with Alexander Gepperth.
• Mathieu Dubois (Post-Doc 2012) : Multimodal semantic segmentation.
• Alexandre Armand (PhD 2011-2016) : Situation understanding and risk assessment framework for preventive driver assistance, Thèse CIFRE Renault, Co-supervised with Javier-Ibanez Guzman.
• Guillaume Duceux (PhD 2011-2015) : Towards Unsupervised And Incremental Semantic Mapping
• Natalia Lyubova (PhD 2010-2013) : Developmental approach of perception for a humanoid robot.
• Islem Jebari (PhD 2009-2012) : 3D semantic mapping for indoor environment exploration.
• Alexandre Chapoulie (PhD 2009-2012) : Contributions to visual loop closure detection and environment topological segmentation methods. Co-supervised with Patrick Rives.
• Stéphane Bazeille (Post-Doc 2009-2011) : Appearance based topological navigation.
• Adrien Angeli (PhD 2005-2008) : Visual SLAM applications of loop-closure detection. Supervised by Jean-Arcady Meyer.

Former Projects

• H2020 VeriDREAM (2020-2022) The VeriDREAM project (VERtical Innovation in the Domain of Robotics Enabled by Artificial intelligence Mandhods) is a European project with the objective of developing industrial applications following the H2020 DREAM and RobDREAM projects.

• PIA EVAPS (2016-2020) The objective of the EVAPS project (Eco mobility by Autonomous Vehicles in the Paris-Saclay region) is to develop intelligent mobility services for peri-urban journeys, in autonomous driving, without driver. We have developed trajectory planning and tele-operation algorithms for remote vehicle handover.

• ITEA3 DANGUN (2016-2019) Project on the development of an autonomous vehicle in which we have developed a remote operation capacity for these vehicles.

• H2020 DREAM (2015-2018) The DREAM project (Deferred Restructuring of Experience in Autonomous Machines) is a European project focusing on the development of processes inspired by sleep and dreams in a cognitive architecture for robotics. We have developed state representation learning approaches for reinforcement learning.

• PSPC ROMEO 2 (2012-2017) The ROMEO 2 project carried by Aldebaran Robotics in which we have developed object dandection mandhods by RGB-D camera and semantic mapping mandhods.

• FUI ROBOT POPULI (2012-2014) Project with the company Awabot in which we have developed a robust visual navigation algorithm for service robotics.

• ANR MACSi (2010-2014) The MACSi project focuses on the development of social and developmental learning capacities for perception on an iCub robot.

• ANR PACOM (2009-2012) The PACOM project concerns the development of exploration mandhods and semantic mapping within the framework of the “Carrote” compandition organized by the ANR and the DGA.

All my publications, including preprints are available on my google scholar page

Publications accepted at workshops, conferences or journals (from HAL):