image - 如何可视化词袋码本（图像分类）？

Question

我想在图像分类中使用词袋特征，如何可视化码本？我使用 keypoint-sift 然后 kmeans 进行聚类。例如， http: //fias.uni-frankfurt.de/~triesch/courses/260object/papers/Fei-Fei_CVPR2005.pdf（图 4）

Answer 1

The 174-word codebook is visualized by a patch. The paper mentioned they used two types of representation, one is 11*11 pixel patch and the other is SIFT descriptors. Fig 4 is a result based on the former presentation after a k-means clustering. They cannot visualize a codebook based on SIFT (a wired image as 174*128 ). Of course, we can get the closest SIFT in query and visualize a patch around corresponding keypoint. Hope it helps.