Noël Meriaux

Contact Email : noel.meriaux@inria.fr Adresse : Irisa / Inria Rennes Campus de Beaulieu 35 042 Rennes cedex - France Tel : +33 2 99 84 22 46 Fax : +33 2 99 84 XX XX Assistant : +33 2 99 84 22 52 (Hélène de La Ruée)

Background and Position

Diploma of computer engineering, and Master 2 (for research) in image-processing and computer vision.

Right after my high-school graduation exam (2008), in sciences with reinforcement in maths, I accepted the challenge of entering a French "classe préparatoire aux grandes écoles" for two years in the maths and physics department, with computer science as an optional subject. After two intensive years in the Lycée Thiers of Marseille, I was accepted in the ISIMA, a computer engineering and applied mathematics school in Clermont-Ferrand. During my last years, I followed a complementary course to aim at validating a Master as well.

Each year, I made projects in the context of my course, always related to my field of expertise.

• "TIPE" in classe préparatoire : optimisation of the distribution of the surfaces in a Voronoi diagram
• 1^st year, engineering school : implementation in C of several procedures to build up fractal figures
• 2^nd year, engineering school : analyse and modeling in MatLab of vector stochastic processes
• 3^rd year, engineering school : parametric modeling in MatLab of twisted trees

My working experience up to now already forms a ground of valuable knowledge in image processing and computer vision.

• 2^nd year, engineering school : algorithms for video stabilisation and visual tracking (Effidence, Clermont-Fd)
• 3^rd year, engineering school : detection and tracking of vehicles for driving assistance (ITS, Kumamoto Japan)
• 1-year contract, 2013-2014 : sensor data fusion for automatic tracking of a group of persons by a robot (Institut Pascal, Clermont-Fd)
• Short teaching experience in mathematics, physics et chemistry at different levels of secondary schools, in a class or for one particular student

Research areas

The ANR-funded project Visioland let by Airbus aims at finding new visual-based methods to have a plane land automatically. My major tasks are split into two main axes:

• An approach based on a 3D model of the target
• An approach based on dense homography calculation

The main objective is to estimate the pose of the aircraft at each new image, relatively to the world and to the objective, in order to make it follow an ideal trajectory.

Publications

Complete list (with postscript or pdf files if available)

Software

Teaching