CEA LIST (CEA) - France

Website: http://www-list.cea.fr/

The Image Embedded Computer Lab (LCEI) of the CEA LIST employs than 30 members working on vision systems for video surveillance, multimedia applications and 3D vision.

The LCEI has been working for over 15 years on image processing techniques. It has now a strong expertise in object detection, pattern recognition and tracking techniques. It has also participated in many national and European projects (SAVE-U, IST – CLOVIS, Eureka - ISCAPS, PASR). Powerful pattern recognition techniques (boosting, cascade of classifiers…) have been developed for several years and present a real interest for video understanding and semantic, indexing.

Role in the project

The content analysis of video is the central research axis of the LCEI. Thanks to the LINDO project, LCEI wants to adapt and improve its existing techniques such as learning methods to index video in real-time with semantic information and therefore permit to have a first-class storage.

In the LINDO project, it will bring its expertise and technologies in video content analysis. The main contribution will be to provide semantic information on video for a better indexing. For that goal, techniques based on machine learning will be adapted and specialized to give high-level information for a better databases access.