Bio

Research

After having worked on programming languages, software engineering, security, and natural language processing, I have now turned to computer vision: I have joined the IMAGINE group in December 2009. I am interested in the reconstruction of 3D models from images and range data, regarding both geometry and semantics, in particular with applications to building and city modeling.

• -- Camera registration. -- In particular, I have been working on camera calibration issues (external), with a focus on accuracy and robustness, developing adaptive parameterless methods and global registration techniques, based on points matches as well as line segments. This also involves robust feature matching.
• -- Geometry processing. -- I have addressed speed and robustness issues for the treatment of point clouds and I have proposed a relevant method to reconstruct simple watertight piecewise-planar polygonal meshes.
• -- Semantization. -- I have been working on grammar-based approaches to semantically segment 2D and 3D data, proposing not only top-down but also efficient bottom-up methods, relying not only on high-level handwritten grammars but also on grammars that are automatically learned. I also considered less constrained frameworks such as MRF and transition tables, and auto-context.

A new topic, currently at work, concerns robotic applications for civil engineering.