Fig. 12From: Efficient video collection association using geometry-aware Bag-of-Iconics representationsExample of 3D model alignment. Visualizations obtained from the Berlin and London YouTube dataset. Reichastag (a, c) and Brandenburg Gate (b, d) are aligned by video trajectory (f) as shown in (d, e). Tower of London (g, i) and Tower Bridge (h, j) are aligned by the video trajectory (m) as shown in (k, l). Best view in colorBack to article page