Multimedia Research Lab
The Multimedia Research Lab (MRL) develops classification and fusion algorithms to detect buried explosive devices, such as landmines and improvised explosive devices (IED). The MRL develops algorithms for vehicle mounted, hand-held, and robotic systems. The systems include ground penetrating radar and metal detector sensors. The MRL has secured research funding from NSF, US Army, ONR, ARO, and KSEF. Our work, in close collaboration with the countermine division (US Army Night Vision Lab) and the industry (NITEK Inc. and BAE Systems), has resulted in the development of software systems that have been integrated in several detection systems. Some of the systems are being field tested by the government. Other systems have been deployed in hostile regions of the world for both humanitarian and military applications. The MRL also develops algorithms to represent, analyze, visualize, and mine large multimedia (text, image, and video) databases. Specific tasks that are addressed under this area include: (i) image segmentation; (ii) efficient feature extraction and resprentation for text, image, and audio; (iii) mining multi-modal data and learning inter-modality associations; (iv) automated image annotation; (v) content-based image and video retrieval; and (vi) multi-modal information fusion.
Dr. Hichem Frigui, Director