Laurent Perrinet is researcher in Computational Neuroscience at the "Institut de Neurosciences de la Timone" (France), a joint research unit (CNRS / Aix-Marseille Université). He graduated from the aeronautics engineering school SUPAERO, in Toulouse (France) with a signal processing and stochastic modelization degree. He received a PhD. in Cognitive Science in 2003. The thesis was directed by Manuel Samuelides and co-directed by Simon Thorpe. It emphasized on mathematical analysis on temporal spike coding and specifically rank-order coding. This work was extended on multi-scale and adaptive representation of natural static scenes. His research program is now focusing in bridging the complex dynamics of realistic models of large-scale models of spiking neurons with functional models of low-level vision. 1) We study the macroscopic effect of microstructures (di-synaptic thalamo-cortical contacts, patchy connections) by implementing realistic models of the primary visual cortex. These results are confronted to experimental optical imaging data collected conjointly in the lab. 2) We model low-level vision function (tracking or catching an object) by using Bayesian-type probabilistic representations of information. These models introduce non-linearities (contrast gain control, center-surround interactions, adaptation) which are emerging naturally from the simple operations on probabilities. Current challenge is to be able to translate, or compile in computer terminology, such functional models into neural architectures that would exhibit similar dynamics.
Under the supervision of Guillaume Masson at the Mediterranean Institut of Cognitive Neuroscience (INCM-UMR 6193, CNRS) in Marseille, I study models of spatio-temporal inference in low-level sensory areas.
Task-leader of a "work-package in the European project FACETS : organization of 2 workshops (« Coordination on WP9T2: Investigate low-level visual processing in a computational model of primary visual cortex », 2006 and 2007), animation of collaborations. Organization of a practical workshop (« Code-Jam »).
- Teaching : Master students, 1 PhD (Jens Kremkow, defended in 2010), 1 invited PhD (Sylvain Fischer, defended in 2008), 2 post docs.
Member of the "NeuroComp" network:
- Reviewer in different journals and conferences, among which: Journal of Vision, Neurocomputing, ESANN, IEEE Trans. in Image Processing, Biological Cybernetics.
Participation to wide audience scientific events: invited conference "Peut-on parler d’intelligence mécanique?" in 2010, artistic collaboration with Etienne Rey in 2010, publication tqrgetting school audience on computational paradigms in Doc-Sciences (nov' 2010).
10/2004 visiting collaboration @ Bruno Olshausen's lab, working on the Role of homeostasis in learning sparse representations.
10/2010 - 01/2012 collaboration @ Karl Friston's laboratory, on understanding ocular movements using the free energy principle.
Dissertation Topic: How to decipher the visual spike code? Study of the parallel, asynchronous and sparse dynamics in ultra-rapid visual processing.
- Department of Computer and Modeling Sciences, grant form the French Research Department.
- Advisor: Manuel Samuelides, Co-advisor: Simon Thorpe.
Presentations at international meetings (IJCNN99, NeuroColt00, DYNN00, CNS00, CNS01, LFTNC01, ESANN02, NSI02). Participation to the EU Advanced Course in Computational Neuroscience in Trieste (Italie) and to the Telluride Neuromorphic Workshop in Colorado (USA).
- Defended 02/2003 with "Highest Honors" distinction.
Master in Cognitive Science Univ. Paris VII, P. Sabatier, EHESS, Polytechnique (Paris, France). 9/1998-9/1999
- Internship at ONERA/DTIM (Toulouse, France), specialization in modeling and mathematics. Study of hebbian learning in a rank-order based neural network. Application to texture recognition. "Very Good" distinction.
7/1999-8/1999 Research Scholar, USAFB (Rome, NY) / University of San Diego in California (États-Unis).
M.S. in Engineering Supaéro (Toulouse, France), one of the leading French Engineering Schools ("Grandes Ecoles"). 1993 - 1998
Specialization in stochastic models for signal and image processing and particularly by means of artificial neural networks.
- M.S. in Engineering, SUPAERO, Toulouse, France. 1993 - 1998
4/1998-9/1998 Research Scholar, CerCo (CNRS, UMR5549), Toulouse (stage de fin d'études d'ingénieur).
- Implementation and mathematical study of a unsupervised asynchronous neural net and application to pattern and character recognition.
- Programmed on a imaging processing facility involving radar decoding, geographical corrections, filtering and geocoding. Programmed on a topographic map generator. Collaborated in the processing of images designed to monitor the terrestrial crust's deformation.
- Monitoring of recent earthquake in California and in China.
9/1995-6/1996 Engineer at Alcatel, Vienna (Austria). Department of Voice Processing Systems. This voluntary internship involved all stages of the product process: design phase, prototyping and quality control with the objective of producing a high technology Voice Mail System.
- Individual projects on visual perception and object recognition.