Nicolas Perrin-Gilbert
49
Documents
Présentation
Je suis roboticien et chercheur en apprentissage automatique.
Je travaille comme chargé de recherche au CNRS à l'ISIR (https://www.isir.upmc.fr/?lang=en), un laboratoire de Sorbonne Université à Paris, France.
Mots clés : *apprentissage automatique, apprentissage par renforcement, apprentissage de représentation, planification de mouvement, contrôle, locomotion à pattes, robots à pattes*.
Liens : [ISIR](https://www.isir.upmc.fr/personnel/perrin/ "ISIR") / [Github](https://github.com/perrin-isir "Github") / [DBLP](https://dblp.org/pid/37/1452.html "DBLP") / [Google Scholar](https://scholar.google.com/citations?user=_UceE6YAAAAJ&hl=fr "Google Scholar") / [Videos](https://www.youtube.com/channel/UC9incxmKuvpvWWfCIQbQq1A/videos "Videos")
I am a roboticist and machine learning researcher.
I work as a permanent CNRS research scientist (*chargé de recherche*) at [ISIR](https://www.isir.upmc.fr/?lang=en), a laboratory of Sorbonne University in Paris, France.
Key words: *machine learning, reinforcement learning, representation learning, motion planning, control, legged locomotion, legged robots*.
Links: [ISIR](https://www.isir.upmc.fr/personnel/perrin/ "ISIR") / [Github](https://github.com/perrin-isir "Github") / [DBLP](https://dblp.org/pid/37/1452.html "DBLP") / [Google Scholar](https://scholar.google.com/citations?user=_UceE6YAAAAJ&hl=fr "Google Scholar") / [Videos](https://www.youtube.com/channel/UC9incxmKuvpvWWfCIQbQq1A/videos "Videos")
Publications
- 13
- 4
- 1
- 1
- 2
- 6
- 3
- 5
- 3
- 3
- 2
- 2
- 2
- 2
- 4
- 5
- 4
- 3
- 1
- 1
- 1
Layered Controller Synthesis for Dynamic Multi-agent SystemsInternational Conference on Formal Modeling and Analysis of Timed Systems (FORMATS 2023), Sep 2023, Antwerp, Belgium. pp.50-68, ⟨10.1007/978-3-031-42626-1_4⟩
Communication dans un congrès
hal-04237998v1
|
|
The Quality-Diversity Transformer: Generating Behavior-Conditioned Trajectories with Decision TransformersGECCO '23: Genetic and Evolutionary Computation Conference, Jul 2023, Lisbon Portugal, France. pp.1221-1229, ⟨10.1145/3583131.3590433⟩
Communication dans un congrès
hal-04237993v1
|
|
Assessing Quality-Diversity Neuro-Evolution Algorithms Performance in Hard Exploration ProblemsGECCO 2022 Workshop on Quality Diversity Algorithm Benchmarks, Jul 2022, Boston (MA), United States
Communication dans un congrès
hal-04237990v1
|
|
|
DIVERSITY POLICY GRADIENT FOR SAMPLE EFFI-CIENT QUALITY-DIVERSITY OPTIMIZATIONWorkshop on Agent Learning in Open-Endedness (ALOE) at ICLR 2022, 2022, virtual, Vatican City
Communication dans un congrès
hal-03753541v1
|
|
Divide & Conquer Imitation Learning2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022), Oct 2022, Kyoto, Japan. pp.8630-8637, ⟨10.1109/IROS47612.2022.9982020⟩
Communication dans un congrès
hal-03753530v1
|
Diversity policy gradient for sample efficient quality-diversity optimizationGECCO '22: Genetic and Evolutionary Computation Conference, Jul 2022, Boston, United States. pp.1075-1083, ⟨10.1145/3512290.3528845⟩
Communication dans un congrès
hal-03864262v1
|
|
Selection-Expansion: A Unifying Framework for Motion-Planning and Diversity Search Algorithms30th International Conference on Artificial Neural Networks - ICANN 2021, Sep 2021, Bratislava, Slovakia. pp.568-579, ⟨10.1007/978-3-030-86380-7_46⟩
Communication dans un congrès
hal-03404366v1
|
|
First-Order and Second-Order Variants of the Gradient Descent in a Unified Framework30th International Conference on Artificial Neural Networks - ICANN 2021, Sep 2021, Bratislava, Slovakia. pp.197-208, ⟨10.1007/978-3-030-86340-1_16⟩
Communication dans un congrès
hal-03404369v1
|
|
State Representation Learning from Demonstration6th International Conference on Machine Learning, Optimization, and Data Science, LOD 2020, Jul 2020, Siena, Italy
Communication dans un congrès
hal-03083156v1
|
|
PBCS: Efficient Exploration and Exploitation Using a Synergy Between Reinforcement Learning and Motion PlanningArtificial Neural Networks and Machine Learning – ICANN 2020, Sep 2020, Bratislava, Slovakia. pp.295-307, ⟨10.1007/978-3-030-61616-8_24⟩
Communication dans un congrès
hal-03080918v1
|
|
Understanding Failures of Deterministic Actor-Critic with Continuous Action Spaces and Sparse RewardsArtificial Neural Networks and Machine Learning – ICANN 2020, Sep 2020, Bratislava, Slovakia. pp.308-320, ⟨10.1007/978-3-030-61616-8_25⟩
Communication dans un congrès
hal-03080925v1
|
|
Learning Compositional Neural Programs with Recursive Tree Search and PlanningAdvances in Neural Information Processing Systems 32 (NeurIPS 2019), Dec 2019, Vancouver, Canada
Communication dans un congrès
hal-03080949v1
|
|
|
Computing regions of stabilizability for nonlinear control systems with input constraintsAmerican Control Conference (ACC'18), 2018, Milwaukee, United States. pp.2869-2876, ⟨10.23919/ACC.2018.8430798⟩
Communication dans un congrès
hal-03135362v1
|
|
Effective Generation of Dynamically Balanced Locomotion with Multiple Non-coplanar ContactsInternational Symposium on Robotics Research, Sep 2015, Sestri Levante, Italy
Communication dans un congrès
hal-01388632v1
|
|
Dynamically transitioning between surfaces of varying inclinations to achieve uneven-terrain walking2014 IEEE International Conference on Robotics and Automation (ICRA), May 2014, Hong Kong, China. pp.210 - 215, ⟨10.1109/ICRA.2014.6906611⟩
Communication dans un congrès
hal-01394896v1
|
|
Lyapunov Stability Margins for humanoid robot balancing2014 IEEE/RSJ International Conference on Intelligent Robots and Systems, Sep 2014, Chicago, IL, United States. ⟨10.1109/IROS.2014.6942673⟩
Communication dans un congrès
hal-01394903v1
|
|
Compliant attitude control and stepping strategy for balance recovery with the humanoid COMAN2013 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2013), Nov 2013, Tokyo, Japan. pp.4145-4151, ⟨10.1109/IROS.2013.6696950⟩
Communication dans un congrès
hal-03089490v1
|
|
Upper limb compliant strategy exploiting external physical constraints for humanoid fall avoidance2013 13th IEEE-RAS International Conference on Humanoid Robots (Humanoids 2013), Oct 2013, Atlanta, United States. pp.397-402, ⟨10.1109/HUMANOIDS.2013.7030005⟩
Communication dans un congrès
hal-03089494v1
|
|
From Discrete to Continuous Motion PlanningTenth Workshop on the Algorithmic Foundations of Robotics, Jun 2012, Cambridge, Massachusetts, United States. pp.89-104, ⟨10.1007/978-3-642-36279-8_6⟩
Communication dans un congrès
hal-03089485v1
|
|
Real-time footstep planning for humanoid robots among 3D obstacles using a hybrid bounding boxIEEE International Conference on Robotics and Automation, May 2012, Saint Paul, Minnesota, United States. pp.0-0
Communication dans un congrès
hal-00668794v1
|
|
LMPC based online generation of more efficient walking motionsIEEE-RAS International Conference on Humanoid Robots (Humanoids), Nov 2012, Osaka, Japan. pp.390-395, ⟨10.1109/HUMANOIDS.2012.6651549⟩
Communication dans un congrès
hal-02487630v1
|
|
A Biped Walking Pattern Generator based on "Half-Steps" for Dimensionality ReductionIEEE International Conference on Robotics and Automation (ICRA 2011), May 2011, Shanghai, China
Communication dans un congrès
hal-00593659v1
|
|
Weakly collision-free paths for continuous humanoid footstep planningInternational Conference on Intelligent Robots and Systems (IROS), 2011 IEEE/RSJ, Sep 2011, San Francisco, United States. pp.4408 -4413, ⟨10.1109/IROS.2011.6048120⟩
Communication dans un congrès
hal-00639026v1
|
|
Real-time Replanning Using 3D Environment for Humanoid RobotIEEE-RAS International Conference on Humanoid Robots (HUMANOIDS 2011), Oct 2011, Bled, Slovenia. p.584-589
Communication dans un congrès
hal-00601300v1
|
|
Walking without thinking about itIROS 2010 - IEEE-RSJ International Conference on Intelligent Robots & Systems, Oct 2010, Taipei, Taiwan. pp.190-195, ⟨10.1109/IROS.2010.5654429⟩
Communication dans un congrès
inria-00567668v1
|
|
Approximation of Feasibility Tests for Reactive Walk on HRP-2IEEE International Conference on Robotics and Automation, May 2010, Anchorage, United States. pp.1-6
Communication dans un congrès
hal-00455454v1
|
|
Fast foot prints re-planning and motion generation during walking in physical human-humanoid interactionHumanoids, Dec 2009, Paris, France. pp.284-289, ⟨10.1109/ICHR.2009.5379563⟩
Communication dans un congrès
hal-00507141v1
|
|
Tree Automata with Memory, Visibility and Structural Constraints10th International Conference on Foundations of Software Science and Computation Structures (FOSSACS), Mar 2007, Braga, Portugal. pp.168-182, ⟨10.1007/978-3-540-71389-0_13⟩
Communication dans un congrès
inria-00579009v1
|
|
Biped Footstep PlanningAmbarish Goswami; Prahlad Vadakkepat. Humanoid Robotics: A Reference, Springer Netherlands, pp.1-21, 2017, 978-9400760455. ⟨10.1007/978-94-007-7194-9_29-1⟩
Chapitre d'ouvrage
hal-03089498v1
|
|
Timed-Automata Abstraction of Switched Dynamical Systems Using Control FunnelsFormal Modeling and Analysis of Timed Systems, 9268, Springer International Publishing, pp.60 - 75, 2015, Lecture Notes in Computer Science, 978-3-319-22974-4. ⟨10.1007/978-3-319-22975-1_5⟩
Chapitre d'ouvrage
hal-01395104v1
|
|
Selection-Expansion: A Unifying Framework for Motion-Planning and Diversity Search Algorithms2021
Pré-publication, Document de travail
hal-03196479v1
|
Learning Compositional Neural Programs for Continuous Control2020
Pré-publication, Document de travail
hal-03083161v1
|
|
QD-RL: Efficient Mixing of Quality and Diversity in Reinforcement Learning2020
Pré-publication, Document de travail
hal-03083159v1
|
|
Importance mixing: Improving sample reuse in evolutionary policy search methods2019
Pré-publication, Document de travail
hal-02397754v1
|
|
First-order and second-order variants of the gradient descent in a unified framework2019
Pré-publication, Document de travail
hal-02397757v1
|
|
|
A Continuous Approach to Legged Locomotion Planning2013
Pré-publication, Document de travail
hal-00805981v1
|
|
Adaptive sampling-based approximation of the sign of multivariate real-valued functions2010
Pré-publication, Document de travail
hal-00544891v1
|
|
Warm-Starting Fixed-Point Based Control Synthesis[Research Report] University of Michigan-Ann Arbor; Cnrs; Sorbonne Université. 2018
Rapport
hal-02397867v1
|
|
Footstep planning for humanoid robots : discrete and continuous approachesOther [cs.OH]. Institut National Polytechnique de Toulouse - INPT, 2011. English. ⟨NNT : 2011INPT0080⟩
Thèse
tel-04241437v2
|