Figure 1

Comparing BWs and CBW representations. (a) Representing shapes with one global Bag-of-Words model. The left and the right shapes are both composed with 5 different words: a, b, c, d, e. Both feature vectors are , which count the occurrences of each word. That means using BWs representation, the left and the right shapes are regarded as the same. (b) Representing shapes with Concentric Bags-of-Words model. Even there are the same two shapes as shown in (a), because of the concentric sphere partitioning, the left and the right shapes are different. Along the arrow's direction, counting from the outer sphere to the inner one, their feature vectors are [2 3 1 3 3; 1 1 5 1 1; 0 1 1 0 1] and [0 3 3 2 3; 3 1 2 2 2; 0 1 2 0 0], respectively.