Sites Inria

Version française

Thesis Prize


Hicham Badri, winner of the 2015 AFRIF best thesis prize

Moroccan PhD student Hicham Badri (GeoStat team, Inria - Bordeaux Sud-Ouest) has been awarded 2015 Best Thesis Prize by AFRIF*, the Association Française pour la Reconnaissance et l'Interprétation des Formes (French Association for the Recognition and Interpretation of Shapes). Let's take a look at this prestigious prize and his experience.

Can you explain the subject of your thesis?

My thesis, entitled "Sparse and Scale-Invariant Methods in Image Processing" presents new approaches based on sparsity and scale invariance for the development of quick and effective image processing techniques.  My research enables the development of methods and applications to solve problems linked to image processing - such as image recognition, improving image quality or its 3D reconstruction.

I defended my thesis on 1st December 2015. I prepared and completed it under the joint supervision of professors Hussein Yahia (Inria Bordeaux Sud-Ouest) and Driss Aboutajdine (FSR/CNRST (National Centre for Scientific and Technical Research), University Mohammed V in Rabat).

Where did the idea to apply come from?

It was not my idea at all in the beginning, but that of Josiane Zerubia (Inria - Sophia Antipolis) who was one of my thesis referees. Then Hussein Yahia helped me to put together my application and also supported me in carrying out this project - I would like to thank him as well as the entire GeoStat team.

What did it bring you?

Firstly, it is an honour to receive such a prestigious award. It then enabled me to present the entirety of my work during the RFIA (Shape Recognition and Artificial Intelligence) conference on 29 June last. It was a unique opportunity because, in general, at conferences we only present part of our research.

What are you doing now?

I am working in Berlin for an e-commerce platform. The principle is to redirect users towards the commercial websites that offer the most relevant products with regard to their initial search via the meta-search technique. I am continuing to develop models that enable an improvement in image recognition in order to reference and classify, through images, products from partner sites and to best respond to just-in-time customer requirements.

Thank you for answering our questions and congratulations!