LEAR and XRCE's Participation to Visual Concept Detection Task - ImageCLEF 2010