DCVRA 2015 - Workshop on Distributed and Cooperative Visual Representation and Analysis (DCVRA)

Traditional visual analysis algorithms have been mostly studied in a centralized scenario where all visual data is processed at a central location. However, for emerging applications such as mobile visual search, wireless camera networks and mobile augmented reality, several constraints in computational power, energy and bandwidth require a radically different solution. To enable a paradigm shift from centralized to distributed visual processing, challenges in computational efficiency, feature representation, energy consumption, data compression, object detection and tracking must be addressed. In addition, since the scale of visual data is massive, efficient representation methods and collaboration among the distributed entities are necessary to achieve rapid visual processing, largescale storage and flexible/scalable visual analysis.