Alban Laflaquière
Robotics and AI Researcher
About me
I am a senior research scientist at SoftBank Robotics Europe in Paris, where I am also leading a fundamental research team (AI Lab) dedicated to Developmental Robotics and Machine Learning applied to Robotics.
Previously, I completed my Ph.D. in Robotics at ISIR in Sorbonne University, under the supervision of Bruno Gas, where I also frequently worked with J.Kevin O'Regan. I received my Master's degree in Intelligent Systems from Sorbonne University, where I started working on research in Artificial Perception. Before that I received my engineering degree from ENSEEIHT, during which I spent an exchange year in EPFL and worked on Bio-Inspired and Evolutionary Robotics in the LIS led by Dario Floreano.
Research
My research interests are in Machine Learning and Robotics. I am particularly interested in the emergence and development of intelligence in artificial systems. My goal is to make robots more autonomous in their ability to explore and learn from their environment, taking inspiration from the development of intelligence in animals and humans.
My work focuses on the unsupervised grounding of perceptive abilities and common sense knowledge. Besides the acquisition of functional perception, I am especially interested in the sensorimotor structure underlying perceptive concepts (space, objects, body, modalities) that could act as building blocks for advanced cognitive abilities.
I also keep a close eye on Evolutionary Algorithms as a potential way to design suitable priors for Robotic systems and bootstrap their cognitive development.
Selected Publications
Unsupervised Learning and Exploration of Reachable Outcome Space, G. Paolo, A. Laflaquiere, A. Coninx, S. Doncieux , ICRA 2020 [paper]
Novelty Search makes Evolvability Inevitable, S Doncieux, G. Paolo, A. Laflaquière, A. Coninx, GECCO 2020 [paper]
Unsupervised Emergence of Egocentric Spatial Structure from Sensorimotor Prediction, A. Laflaquière, M. Garcia Ortiz, NeurIPS 2019 [paper]
Self-Supervised Body Image Acquisition Using a Deep Neural Network for Sensorimotor Prediction, A. Laflaquière, V. Hafner, ICDL-Epirob 2019 [paper]
Novelty Search: a Theoretical Perspective, S. Doncieux, A. Laflaquière, A. Coninx, GECCO 2019 [paper]
A Bi-directional Multiple Timescales LSTM Model for Grounding of Actions and Verbs, A. Antunes, A. Laflaquière, T. Ogata, A. Cangelosi, IROS 2019 [paper]
Online Learning of Body Orientation Control on a Humanoid Robot using Finite Element Goal Babbling, P. Loviken, N. Hemion, A. Laflaquiere, M. Spranger, A. Cangelosi, IROS 2018 [paper]
Learning representations of spatial displacement through sensorimotor prediction, M. Garcia Ortiz, A. Laflaquière, ICDL-Epriob 2018 [paper]
A Sensorimotor Perspective on Grounding the Semantic of Simple Visual Features, A. Laflaquière, ICDL-Epirob 2018 [paper]
Discovering space—Grounding spatial topology and metric regularity in a naive agent’s sensorimotor experience, A. Laflaquière, J.K. O’Regan, B. Gas, A.V. Terekhov, Neural Networks, 2018 [paper]
Identification of invariant sensorimotor structures as a prerequisite for the discovery of objects, N. Le Hir, O. Sigaud, A. Laflaquière, Frontiers Robotics and AI, 2018 [paper]
Grounding the experience of a visual field through sensorimotor contingencies, A. Laflaquière, Neurocomputing, 2017 [paper]
Learning agent’s spatial configuration from sensorimotor invariants, A. Laflaquière, J.K. O’Regan, S. Argentieri, B. Gas, A.V. Terekhov, Robotics and Autonomous Systems, 2015 [paper]
Grounding object perception in a naive agent's sensorimotor experience, A. Laflaquiere, N. Hemion, ICDL-Epirob 2015 [paper]
Learning an internal representation of the end-effector configuration space, A. Laflaquiere, A.V. Terekhov, B. Gas, J.K. O'Regan, IROS 2013 [paper]
A non-linear approach to space dimension perception by a naive agent, A. Laflaquiere, S. Argentieri, O. Breysse, S. Genet, B. Gas, IROS 2012 [paper]
Invited Talks
Mapping the self: infants, robots, and modeling, ICDL-Epirob: Discovering space: Knowing where my body is moving, August 2019
SorBotic conference: Formations, prospects and stakes in Robotics, November 2019
Learning Body Models seminar, Lorentz Center: Robots learning about space and a robot manufacturer perspective, October 2018
Exobiology Meetings, RED: Robotics and AI, Mars 2018
Intervention in the “Excellency Interviews”, May 2011
Extra
Emergence of Spatial Coordinates via Exploration, Alban Laflaquière
Work presented at BabyMind @NeurIPS 2020
Nao robot learns to control its orientation in 25 minutes, Pontus Loviken, Nikolas Hemion, Alban Laflaquière
Nao learns to body control from scratch blog post
Obstacle Detection with the Pepper Robot, Abdelhak Loukkal, Alban Laflaquière, 2017
Unsupervised Learning and Exploration of Reachable Outcome Space, Giuseppe Paolo, Alban Laflaquière, Alexandre Coninx, Stéphane Doncieux, 2019
Self-Supervised Prediction of Body Image, A Laflaquière, VV Hafner, 2019
Robotics and AI Talk at RED exbiology 2018
Associative life
Member of the Meta-Orchestra, innovative music project in Puce-Muse, 2009 - 2013
Organizer and participant in the French Science Week, 2008 - 2012
Main coordinator and organizer of the Day for the Young Researchers in Robotics (JJCR), 2010
President of the Ultimate Sport Association, ENSEEIHT, 2006 - 2007