Vajda, PéterIvanov, IvanGoldmann, LutzEbrahimi, Touradj2010-10-222010-10-222010-10-22201110.1109/DSP-SPE.2011.5739235https://infoscience.epfl.ch/handle/20.500.14299/56030WOS:000298803400059With the technological evolution of digital acquisition and content analysis, millions of images and video sequences are captured every day and used in a large variety of applications. As keyword-based indexing is very time consuming and inefficient due to linguistic and semantic ambiguities, content-based image and video retrieval systems have been proposed, which search and retrieve documents based on the content itself rather than its associated tags or keywords. Within such systems, a query document is usually compared to all the documents in a database through visual features extracted from it. However, since the features are extracted from images which contain two-dimensional projections of three-dimensional objects, the features may change significantly depending on the view point. Thus, systems could fail to retrieve relevant content in response to some queries.enobject duplicate detectiongraph matchingSIFTomnidirectional detectionvisual searchtext::conference output::conference proceedings::conference paper