Renaud Marlet is currently a Senior Researcher at the IMAGINE/LIGM lab, École des Ponts ParisTech (ENPC), Marne-la-Vallée, France. He has various fields of interest, including programming languages and software engineering, security, computational linguistics, and computer vision, and has worked in both academia and industry.

After working on programming languages, software engineering, security, and natural language processing, he has now turned to computer vision: he joined the IMAGINE group in December 2009. He is interested in reconstructing 3D models from images and range data, addressing both geometry and semantics, particularly with applications in building and city modeling.

Camera registration. In particular, he has worked on camera (external) calibration problems, focusing on accuracy and robustness, developing parameter-free adaptive methods and global registration techniques based on both point correspondences and line segments. This also involves robust feature matching.

Geometric processing. He has tackled speed and robustness issues for point cloud processing and proposed a relevant method to reconstruct simple watertight polygonal meshes piecewise.

Semantization. He has worked on grammar-based approaches to semantically segment 2D and 3D data, proposing not only top-down methods but also efficient bottom-up methods, relying not only on high-level handcrafted grammars but also on grammars that are automatically learned. He has also considered less constrained frameworks such as MRFs and transition tables, as well as automatic context.

A new subject, currently ongoing, concerns robotic applications for civil engineering.

Computer Vision, Scene Understanding, 3D, Geometry Processing