Ingénieur de Recherche Contractuel CNRS à Grenoble, France:

I am currently employed as a Research Engineer for the Camomile project ( http://camomile.limsi.fr), a french collaborative research project, which aims at creating a collaborative annotation framework for 3M (multimodal, multimedia and multilingual) data. My main task within the project focuses on Information Indexing, Information Retrieval & Corpus Annotation. Mainly, to apply an active learning system that directs human anotators to label the existence of given personnes within the video corpus.

My research focuses on three main areas: Automatic Image and Video Indexing, Information Retrieval, Annotations by Active Learning & Active Cleaning, Machine Learning and design and management of data in order to build a powerfull applications for visual concept detection and retrieval in multimedia documents. My specific interests are in semantic indexing of visual concepts in very large-scale multimedia corpuses and the use of active learning approaches to build annotated corpus. Semantic indexing is so far the most needed functionality for users, however, it is also the most difficult one, due to the semantic-gap and class-imbalance problems. This is motivated by the need to provide powerful tools to extract the correct content from multimedia documents. For corpus annotation, the main objective of using active learning is to increase the system performance by using as few labeled samples as possible, thereby minimizing the cost of labeling data (e.g. money and time).

Additionally, I am also interested on using "Internet-scale" data and crowd-sourcing to improve scene understanding and allow smarter image synthesis and manipulation. Also, I am very interested in Computer Vision researches and their applications to robotics, and the healthcare computing by implementing specific techniques to develop systems for acquisition, processing and interpretation of medical data.