14. K. Mikolajczyk: Efficient Clustering and Matching of Local Features

In this work we address the problem of building object class representations based on local features and fast matching in a large database.
We examine different agglomerative and partitional clustering strategies and compare the quality of obtained clusters.
We also evaluate methods for building data structures and different search strategies in high dimensional feature spaces.
In particular we compare the approaches proposed by Nister-CVPR06 and Mikolajczyk-CVPR06.
These methods allow to deal with large sets of data typically used in image retrieval and multiclass recognition.