• Statistical machine learning

I thus develop computational models, at the interface between cognitive science domains, including computer science, psychology, neuroscience and robotics.

Associated results are illustrated through my publications and the dedicated projects pages. A not-up-to-date manifesto also develops my cognitive science perspective.

Mathematical formalism

• Complex dynamical systems

Experimental approach

• Artificial sensorimotor systems (robots)
• Dynamical measures of performance (human psychology)

Computational approach

• Distributed anticipatory systems (multi-agent like)
• Active perception models
• Network based representations

Applications

• Decision-making
• Active vision
• Navigation/localization
• Manipulation

Associated results are illustrated through my publications and the dedicated projects pages. A not-up-to-date manifesto also develops my cognitive science perspective.

Associated results are illustrated through my publications and the dedicated projects pages. A not-up-to-date manifesto also develops a bit more my cognitive science perspective on these research topics.

Generally speaking, I'm interested in distributed models of life and cognition, a topic which takes different forms depending on the disciplinary perspective adopted. I usually deal with complex dynamical systems endowed with predictive capabilities, embodied in natural (human) or artificial (robotic) systems, maintaining continuous sensorimotor interactions with their environment.

The topics and the associated results are illustrated through my publications and the dedicated projects pages.

Theoretical stance

• Predictive coding
• Interactivism
• Enaction
• Embodiment

Mathematical stance

• Dynamic systems formalism
• Complex systems models
• Topological (self-)organization
• Dissimilarity spaces
• Kernel methods

Generally speaking, I'm interested in distributed models of life and cognition, a topic which takes different forms depending on the disciplinary perspective adopted. I usually deal with complex dynamical systems endowed with predictive capabilities, embodied in natural (human) or artificial (robotic) systems, maintaining continuous sensorimotor interactions with their environment.

Generally speaking, I'm interested in distributed models of life and cognition, with a focus on anticipatory systems. In practice, I use computational models to test and refine hypotheses, mainly studying sensorimotor behaviors.

Research keywords:

• Anticipatory systems (for control and perception)
• Active perception (especially in vision)
• Embodied cognition (brain/body/environment coupling)
• Distributed models (bio and neuro-inspired)

Research topics

Since December 2011, I am an assistant professor at Clermont University (France), teaching at Polytech Clermont-Ferrand. This engineering school is composed of 6 departments (electrical, math and modeling, biological, civil, physical, production engineering) and I teach computer science in most of them.

Follows a non exhaustive list of courses I've already taught, including those at my previous institutions (e.g. University of Toulouse / ENSEEIHT engineering school in 2005-2009, University of Lorraine / Nancy 2 in 2009-2010).

Current courses

• Algorithmics (basics, structures) (L3)
• Imperative programming (C/VBA/Pascal) (L3)
• Object oriented programming (UML, C++, Java/C#) (M1)
• Operating systems (Unix/Linux, tools, scripts) (L3)
• Databases (E/A, relational algebra, Oracle SQL, MySQL) (L3)
• Web technologies (network protocols, Internet, HTML5/CSS3) (L3)

Past courses

• Certificate In Computers and Internet (C2I) (LibreOffice, Microsoft Office suites, Web) (L3)
• Web technologies (dynamic website design, XHTML, CSS, PHP, SQL databases, XML/XSL/XSD) (M1)
• Computer architecture (logics, (a)synchronous circuits, processors, FPGA simulation/emulation) (L3)
• Imperative programming (C/Fortran, static/dynamic memory allocation, structures/algorithms) (L3)
• Functional programming (Caml, recursion, typing and polymorphism, higher order functions) (L3)
• Centralized systems (UNIX/Linux, Shell scripts, ssh, core primitives, processes/threads) (L3, M1)
• Data mining and analysis (sampling, quantification, PCA, Bayesian, regression/classification) (M1)
• Natural language processing (Prolog, syntactic/semantic analysis, logic and categorial grammars) (M2)
• Artificial intelligence (decision making/planning, fuzzy/probabilistic logic, multi-agent systems) (M2)
• Image synthesis/analysis (filters/segmentation, 2D/3D transforms, augmented reality, filling...) (M2)
• Geometrical modelling (CAO, OpenGL, modelling, subdivision, rendering pipelines) (M2)

Projects and software development

Interacting with Master level engineering students offers the opportunity to propose long-term projects. It is then a great pleasure to monitor the progress of a student team working full-time in a quasi-professional manner on topics of interest:

• Driving simulator (realistic physics, interaction/feedback for autonomous AI control)
• Haptic device control (human-haptic device interactions for handwriting analysis and rehabilitation)
• Image processing on the Cell BE/PS3 (distributed/pipelined algorithms for HD video stream processing)
• Saliency FPGA implementation (Itti & Koch algorithm implementation on FPGA)
• 2.5D realistic pictures synthesis (combining real photos in fog to reduce experimental cost)
• Dynamic Neural Field on FPGA (optimized kernels, network of processor architecture)
• MouseTracker Java implementation (for flexibility, portability, robustness and performance)
• Automated RAPM (Raven Advanced Progressive Matrices on a computer for dynamical analysis)
• Continuous Tetris implementation (to analyze real-time decision making and planning)
• Visualization of dynamical complex systems (study and rapid prototyping with high dimensionality)
• R package for MouseTracker (multi-level data analysis and visualization)

(softwares and other resources will be put again online ASAP)

Philosophy

My general approach to teaching is to propose a variety of supports and methods, as to let students find the most appropriate way to assimilate knowledge (depending on their own perceptual and learning preferences). Particularly committed to a constructivist approach to development, I try to create interactive applications whenever possible, as to let students explore and learn at their own pace even when at home. To promote such dynamics, they have to be motivated to do so and therefore must be interested in the subject, which is in my opinion the goal of the teacher's interactions with them. This might not be adapted to all situations, but might well be worth the effort when dealing with students that in general all have the capacity to succeed.

(softwares and other resources will be put again online ASAP)