Current Position

I am currently associate professor at the ENSEA, a french grande école for electrical engineering and computer science. I'm part of the ETIS lab, within the MIDI team.

Contact:
David Picard
picard (at) ensea [dot] fr
Office 312
6 avenue du Ponceau
95014 Cergy-Pontoise
France

Research Interests

My actual research interests include Computer Vision and Machine Learning for Information Retrieval, Content Based Image Retrieval and Multimedia Processing. In this topic, I focus on Kernel Based algorithms, with the goal of defining new kernels that better encapsulate the property of images (spatial or contextual information, for instance). This new types of kernels can be learned from labeled samples or via user interaction. The other big part of my research is focused of distributed retrieval, especially for multimedia content. Today's multimedia content are highly distributed (think of the Internet), and my work in this area is to adapt information retrieval algorithms so as to use this distribution as an additionnal information source.

Education

I received the M.Sc. in Electrical Engineering degree from the ENSEA (Cergy-Pontoise - France) in 2005, with a strong emphasis on computer science. The same year, I received the M.Sc. in Image and Signal Processing degree from the Univesité de Cergy-Pontoise.
I successfully defended my Ph.D. in 2008 under the supervision of Matthieu Cord and Arnaud Revel. This work is about Interactive Content Based Image Retrieval in a distributed context with the use of Multi-Agents Systems. I developped a system in which agents crawl the network in order to retrieve images of interest. Agents mark the sites they visit with an ant-like reactive algorithm that speeds up the discovery of usefull sites by other agents. Using these bio-inspired strategies, the system is able to shorten the time taken for the interaction within the relevance feedback loop, while keeping a high rate of relevant results.



Link to CV in french.