Fig. 2From: Efficient video collection association using geometry-aware Bag-of-Iconics representationsVisualization of image clustering on London Flickr dataset (see Section 4.4). First row: iconic views for different connect components. From left to right: Big Ben, Westminster Abbey, London Eye, Buckingham Palace, and Tower Bridge. Second row: selected images from one of the Big Ben image clusters. Images cropped for visualization purposes. Best view in colorBack to article page