Who are you and what are you working on?
I have a PhD in computer science from the Royal Institute of Technology in Stockholm, in Sweden. For over 15 years I have been focusing on the visual recognition of objects, actions and interactions, computer vision and machine learning. I am currently an Inria senior scientist within the project team "Willow: Models of visual object recognition and scene understanding", hosted at the Computer Science department of the École Normale Supérieure I have also been chosen to represent and manage the Louis Vuitton research chair dedicated to artificial intelligence and which was originally put in place by Jean Ponce, one of my colleagues in the Willow team.
In what way is your work of interest to this chair?
The aim of Willow is to develop statistical, geometric and physical models to describe the interpretation process of fixed or animated images. Our research focuses on the visual recognition of objects, but also of movements, actions and interactions between the different elements of the same scene. We are also working on image research, image and language modelling, machine learning of weakly-supervised visual models and 3D scene modelling. Our research findings can prove to be extremely useful in the development of 'intelligent' machines.
What are the interests for Louis Vuitton?
Intelligent data processing is a cornerstone of Louis Vuitton's digital transformation project. With the aim of accelerating this transformation, this world leader in luxury leather goods and fashion has decided to capitalise on the world-renowned scientific expertise of the ENS in the field of artificial intelligence. The creation of the Louis Vuitton chair of excellence makes this ambition a reality. Ultimately, it will enable the brand to have innovative technologies at its disposal in order to globally improve the service it provides to its customers.
What is the interest for your team and for Inria?
With this chair, Louis Vuitton is supporting the research in machine learning and artificial vision carried out in my team and the team "Sierra: Statistical machine learning and parsimony", also hosted at the Computer Science department of the École Normale Supérieure. This collaboration gives us an enormous opportunity - that of tackling concrete industrial problems, testing our algorithms and validating our models in real conditions, and steering our work towards new research opportunities. For Inria, it is also the opportunity to establish strong links with a renowned company and to apply the contributions of its researchers to the scientific challenges it faces in artificial intelligence.