Inria is digitalizing our environment
Pierre Alliez et Mathieu Desbrun (de gauche à droite)
Display our world in three dimensions: that is the objective of the Titane team, directed by Pierre Alliez. For the next five years, Pierre will be working with Professor Mathieu Desbrun, the Head of the Information Sciences and Mathematics Department of Caltech (United States), within the framework of an Inria international Chair.
What is the subject of the Titane team’s research?
Pierre Alliez : Titane is a young team working on 3D geometric modelling of large spaces and complex shapes. If we take measurements using a laser, we are measuring on the scale of a whole town or district. So what we are studying is no longer a shape, but an entire scene. We use discrete measurements, in other words, discontinuous measurements made up of numerous points, then we convert these into a representable scene.
There have been two major developments in our research themes: First, the creation of the Geometrica team in 2003, which marked the switchover from algorithmic geometry to geometrical calculation. The work carried out by this team allowed us to produce algorithms that correspond to the power of the computers in the “real world”, along with their limits. Then, since 2012, we have been looking into processing data from the real world, with all its diversified, heterogeneous and disseminated sensors. The arrival of this imperfect data lowered the quality of the information based on which we work. So we had to adapt our algorithms to this situation. In particular, we created the Titane team in 2013 with the aim of obtaining geometrical computer modelling of complex scenes based on physical measurements
Mathieu Desbrun : For example, look at measurements made by cars equipped with laser sensors: we obtain raw 3D data without any structure. We have to make a geometrical meaning out of this so that we can recognise a street, a tree or a building. My specific mission will be to help the team to increase the reliability of the data with which they are working, to make sure that any false or aberrant measurements will affect it as little as possible.
How is your work useful to industry?
P.A. : It is used through the start-up Geometry Factory, that commercialises our algorithm library CGAL. It also provides a mine of information from the users and this helps us to improve and better target our research
M.D. : As I see it, the purpose of research is to develop new tools for engineers. However, we research scientists always find it complicated to directly address the industrialists, because we have to know exactly what their field of activity is. Geometry Factory does this for us.
For how long have you been working together ?
M.D. : This is our thirteenth year together, with at least one work published per year! After Pierre Alliez’ thesis on digital geometry, I had invited him to do a post-PhD with me at the University of South California in Los Angeles in 2000. This time, it was Pierre who informed me that the Chair was available. I applied for it as soon as the procedure started in 2013, so now it’s my turn to move.
P.A. : Mathieu Desbrun’s team and their works are well known here: we were very enthusiastic at the idea of working with him.
How is your time in France organized?
M.D. : This Chair means that I have to spread 12 months at Inria over five years. I very much like this idea. It is rare and very attractive. Without this flexibility, I would have had to take sabbatical leave from Caltech. During my time in France, I will organize conferences to speak about my works. The first conference already took place in mid-June on the subject of: Power duals for modelling and animation.
Does having the Chair limit research field possibilities?
M.D. : No, the research field is indicated in my application, however changing direction would not be a problem. We have been working together for a long time, so we have a good idea of what can work and what is more risky.
P.A. : What I appreciate the most about international Chairs is that they give the opportunity of looking into complex subjects, making a new start, or reviving subjects that had been in a rut up until now.