Zoom:
Toward Optimal Sampling

Bruno Lévy was the first INRIA researcher to be selected from among 10,000 applicants by the European Research Council, the new scientific funding institution designed to support innovative research in emerging fields.

Interview with Bruno Lévy, head of the ALICE project-team.

Bruno Levy © INRIA / Photo C. Lebedinsky

: What was it about your project that won over the jury?

Bruno Levy: It's a project involving several basic aspects of 3D modeling and computer graphics, which we hope will have interesting outcomes in terms of applications. The field of digital geometry processing has been expanding rapidly for the past ten or so years. Advances have already been made with the transition from digital representations of data to symbolic representations using mathematical functions. Our objective is to progress even further and adapt the concepts of information theory to reformulate problems in terms of digital optimization. The methods that we hope to develop through this approach could eliminate several of the obstacles blocking current methods.

: More specifically, what will you be working on?

Bruno Levy: We are starting out by taking a new approach to the classic, essential problem of sampling, or the digital representation of objects in a computer. The Lloyd-Max algorithm, invented in the 1950s, can be used to calculate optimal sampling, but because it is inefficient (limited to the speed of linear convergence), it cannot be applied to large-scale problems. Our approach will probably result in a new algorithm that is much more efficient in terms of algorithmic complexity and can be applied to a larger class of problems. We will experiment with applications in inverse engineering and oil exploration.

: What does the grant mean to you?

Bruno Levy: First of all, it guarantees that I'll be able to finance four thesis students, two post-doctorates and one engineer using the grant, which totals one million Euros over five years. Even more importantly, however, the ERC's support offers strong encouragement to pursue this difficult and original direction, which represents a risk for my team. Studying these basic problems requires new mathematical concepts, and the fact that our approach has been validated by specialists gives us more confidence to start down the path.

Contact:

Bruno Lévy
ALICE project-team, INRIA Nancy - Grand Est
Tel.: + 33 3 83 59 30 70