Fig. 5. Visualization of high similarity segments for two query shapes: in each part, for every shape of the training set (right), we show with a continuous bold line the high similarity segments after the pairwise alignment with the query shape (left).